by
CLINTON STAPLES
A thesis
presented to the University of Manitoba
in partial fulfillment of the
requirements for the degree of
Master of Arts
rn
History
Whnipeg, Manitoba
Clint Staples, 1.994
WWW NationalLibrary Bibtiothèque narionate
w -r W oluanada duCanada
Acquisitions and Direction des acquisitions et
BibliographicServices Branch des services bibli'ographiques
Street
395 Wellington 39S. rue Weltinoton
Ontario
Otlawa, Ottawa lOntarió
KlA 0N4 xlA o¡là '
y*t t¡to votte fttêterce
ISBN 0-315-98988-2
{4".€.a ^ ÆgåE
\-/d-9. ^B.@q.,Ed-
'F
lfil<, ;,-t J '
<a>(^.¡^ ,
.- r, I \t-
SUBJECT TER"M
.' 2 nz a-
- rtrtril¡t
SUBJECT CODE
KJ&4å
5ubleet €ofegonies
7L5-775 A,.D -
by
GLII{ÏON STAPLES
A Thesis submitted to the Faculty of G¡aduate Studies of the University of Manitoba in partial
fulfillment of the requirenents fo¡ the degree of
MASTER OF ARTS
G) 199{"
Pe¡:nission has been granted to the LIBRARY OF THE LINT\TERSITY OF IvfANTTOBA to lend or
sell copies of this thesis, to the NATIONAL LIBR.ARY OF CA}ÍADA to microfilm this thesis and
to lend or sell copies of the film, and tlNTlrtrSITY ñdICROFILT}ÍS to publish an abstract of this
thesis.
The author reserves other publications rights, and neiiher the thesis nor extensive ext¡acts from it
may be prinæd or otherwise reproduced without the autho/s per:urissíon.
Tabne of Co¡rtemts
Introduction.
Historiography 5
Muslim Sources 22
Leo III
Constantinev... 77
Conclusion L08
Bibliography 115
Acknowledgments
The fortune of the Byzantine Empire over the course of the eighth
continued existence against Muslim forces. The respected Byzantinist J.B. Bury
described the situation as follows: "Asia Minor, however, during the eighth
century was as much exposed as ever to the inroads of the Moslem, who
entered by the Cilician Gates and plundered in one year Cappadocia, and in
In particular, the reigns of the first two Isaurian emperors, Leo Itr and
his son Constantine V, have been seen as a period in which the Empire
suffered extensively, not only from the external threat of the forces of Islam
and of Bulgaria, but also from intense disruption due to the iconoclastic actions
of its rulers.2 This view is largely based upon the testimony of one ninth
1
¡.8. Bury, A History of the Later Roman Empire from Arcadius to lrene.(395-
800). Adolf M. Hakkert. Publ; (Amsterdam,1966) 405; (1st ed. 1889).
2
Louis Brehier states: "The achievements of the Isaurian emperors . . . consisted in
halting the disintegration of the Empire and in protecting it against invasions; but the
task was made difficult and incomplete by the intemal troubles caused by the
iconoclast movement, which brought about the loss of Italy and the West." Life and
Death of B)¡zantium. (Paris, 1946) trans. M. Vaughan; North Holland Publ. Co; New
York, 7977. See also M.V. Anastos, "Iconoclasm and Imperial Rule, 7L7-842,"
century source, the Chronographia, written by the Byzantine monk
Theophanes. Fiercely iconophile, Theophanes must be immediately suspect in
the reigns of the iconoclastic Emperors were as disastrous for the Empire as
events of the eighth century and is often the only Byzantine source to
comment upon a particular matter. This mearu that one often has to take
much of what it says at face value. It is obvious that Theophanes has a strong
bias against the iconoclastic ideology. However, beyond allowing for some
and eighth centuries and provides us with the means to check some of
Theophanes statements, but its utility in this respect is quite limited and one
Cambridge Medieval Histor)¡. c.3, v.4, pt.I, 1966, 61.-1.08; and C. Diehl, Histor)¡ of the
Byzantine Empire. (1st ed. 1919) AMS Press, New York, 1967,53-72.
' See below, p.20-1,, conceming the limitations of the Breviarium as a means for the
corroboration of the statements of Theophanes.
There are several Musiim ch¡oniclers of considerable value in such a
study. Th"y deal with the eighth century, often in more detail than do
Theophanes and Nikephorus, and they have much to say concerning their
Byzantine neighbors. The chronicler and theologian Tabari provides one such
world to his own time (d.923). The geographer and historian Baladhuri (d.893)
in his Kitab Fuhrh al-Buldan deals with the origins of Islam and has much of
interest for Byzantine studies. Other Muslim sources such as Yakubi, ibn
al-Tiotaoa. Masudi. and the anonvmous Khitab al-Uvun and Hudud al-Alam
may also contribute to our understanding of the Empire in the eighth century.
fact, the Empire is only rarely mentioned i. *y other context by the Arabic
affairs for the greater part of the eighth century, drawing upon both Byzantine
present work will attempt to trace the military fortunes in Anatolia, first of Leo
III, and then of his son, Constantine V. ftr the process, an attempt will also be
made to analyse the actions of both father and son for signs of a coherent
to a better understanding of the reigns of the first two haurian emperors and
of this crucial period in the history of the Empire.
II
HISTORIOGRA"HY
Much has been written in the last century concerning Leo Itr and
period of Byzantine history and could hardly have escaped the scrutiny of
historians. Flowever, the focus of much of this scrutiny has centered upon one
issue - the single greatest dogmatic issue of the eighth and early ninth centuries -
Iconoclasm.a While it is certain that both Leo and Constantine were centrally
necessarily follow that Iconoclasm was the only, or even the greatest, concern
which they had to face. It is not the purpose of the present work to concern
itself with Iconoclasm. Here an attempt will be made to address another major
aspect of the reigns of the first two Isaurian Emperors: the Arab Wars.
The military ability of Leo and Constantine has long been recognised by
historians. However, most assessments are rather vague and have rested
a
for a number of works which treat iconoclasm as the major crisis of the eighth
See n.2,
century. See also Ch.V, 'the Iconoclastic Epoch ,717-867,' in A.A. Vasiliev, Histoq¡ of the
Byzantine Empire, University of Wisconsin press, Madison,196'J,,2U-299; and Ch.III, 'The
Age of Iconoclastic Crisis,711-843,'in G. Ostrogorsky, History of the Byzantine State. Basil
Blackwell, Oxford,1956,730-786. More recently there is S. Gero, Byzantine Iconoclasm
During the Reign of l,eo III: with Particular Anention to the Oriental Sources. Corpus
Scriptorum Christianorum Orientalium, Louvain, 7973,n.41; and Byzantine Iconoclasm
During the Reign of Constantine V: with Particular Attention to the Oriental Sources.
Corpus Scriptorum Christianorum Orientalium, Louvain, 1977,n.52. ln addition, there is
a wealth of articles which deal with the Isaurians and the icons.
6
Arabic sources allows for a more detailed analysis of the situation. The majority
of the historians dealing with this subject have either ignored or slighted the
Byzantine lands during the first two centuries of Islam. While his efforts made
record any reference to Arabic raids made by various Arabic authors, he was not
concerned with reporting on the martial endeavors of their enemies, the Romans.
Thus, one might easily conclude that the Arabic writers are silent upon any
military action initated by the Byzantines, and that the only such mentions are to
be found in Byzantine texts - primarily Theophanes. This is not the case, but it
would appear that for the better part of the last one hundred years, historians
5
ln reference to the military situation of Constantine's time, Bury states: "We need not
pursue all the details of the hostilities between the Empire and the Caliphate in the reign
of Constantine V, læo's son and successor. On the whole the Empire was successful."
. l¿ter Roman Emrrire,
Burv, (395-800),406.
6
E.W. Brooks, "The Arabs in Asia Minor (e1,-750) from Arabic Sources," Tournal of
Hellenic Studies, 18 (1898), 182-208; also, "Byzantines and Arabs in the Time of the Eariy
Abbassids,' English Historical Review,ls (1900), 72847; and, "Byzantines and Arabs in
the Time of the Early Abbassids,Il," English Historical Review, 76 (7901),84-92.
who have commented upon the fortunes of the Byzantines against their Muslim
J.B. Bury, writing in 1889, was aware of the Arabic material but appears
still to have relied primarily upon Theophanes and Nikephorus the Patriarch.T
Byzantine material, Bury renders himself unable to assess the military program
of Leo in any detail. For Constantine's wars, he has even less information from
which to draw.8
Charles Diehl viewed all the actions of Leo and Constantine as part of a
grand plan for goverrunent which encompassed political, economic, moral, Legal,
and religious aims. He characterizes both father and son as capable and
far-sighted, but his primary goal was to attempt to place iconoclasm in, what
martial Prowess: "The problem of the Arab struggle, thery was brilliantly solved
7
Bury, The Later Roman Empire.(395-800),408.
I See above, n.5.
n
C. Diehl, Historv of the Bvzantine Emoire, 57.
by Leo III."10 He concludes by stating that Constantine was able: " . . . [to] move
the imperial border farther east along the entire boundary of Asia Minor by
in the well worn path of Bury and Vasiliev, stating that Leo:
" . . . triumphantly preserved the capital through the great Arab siege of 777-718,
and in his later wars beat the inJidel back to the Taurus frontier."l2 Concerning
10
Vasiliev, Historv of the Bvzantine Emoire, 238.
11
ibid,23g.
12
s. Runciman, Byzantine civilization. Edward Arnold publ; London (1933),43.
13
Runciman,212.
-
9
Emperor's lack of fore-sight in allowing Rome and the Empire's titular Itaiian
pointing out that Tabari's historical work is of great value to the Byzantinist.
and Nikephorus.l6
Louis Bréhier's 1946 work Vie et mort de Byzance, saw the importance of
the Isaurian Emperors to be irì.: " . . . halting the disintegration of the Empire, and
in protecting it from invasions."77 Yethe faults both father and son for causing
religious strife: " . . . at a time when every dogmatic quarrel had ceased and
religious peace seemed assuted."l8 lconoclasm, brought into being by Leo and
the Empire in the face of its enemies and ultimately, to bring about the loss of
Milton Anastos believed that Leo and Constantine could only do as they
did concerning iconoclasm. He felt that the Isaurians were acting from the
staunch conviction that the worship of icons endangered the future of the
la
Ostrogorsky,1,49.
ts
ibid, 152.
1ó
ibid, 131.
17
Bréhier,52.
"ibid 52.
i0
Empire. Thus, as devout insÍuments of God's will, they had no choice but to
attempt to correct the error in the faith. Although Anastos did portray
Constantine as an able soldier, his primary purpose was to place the issue of
More recently, Stephen Gero has attempted to shed new and more
favourable light upon the Isaurians by working with the oriental sources.2o He is
are of value. While he met with only limited success in his efforts and certain
the Empire in the ninth century under such rulers as Nicephorus I (802-L1), and
of the tenth century military emperors. Treadgold sees the actions of the
Empress lrene, both during the reign of her son Constantine VI (780-9n and in
her own right (797-802), as being the first sign of recovery for the Empire and the
te
M.V. Anastos,"Iconoclasm and Imperial Rule, TlT-842," Cambridge Medieval
History. Cambridge University Press, Cambridge (1966), c.3, v .4, pt.1,, 61..
2o
S. Gero, Iconoclasm During the Reign of Leo IIL and Iconoclasm During the Reign of
Constantine V. See also "The Legend of Constantine V as Dragon-Slayer," Greek. Roman
and B)¡zantine Studies. 19, (1978),155-159, in which Gero further develops his thesis that
Constantine has been maligned unduly by Theophanes.
:1
W. Treadgold, The Bvzantine Revival. Stanford University Press, Stanford, 1988.
1,1
other concerns at a time when the Empire's future was by no means assured,. In
contrast, he praises the Empress Irene for furning the army over to bureaucrats
and for paying exorbitant bribes to hold the Arabs across the Taurus.
J.F. Haldon has analysed Isaurian military fortunes under Leo and
Constantine with specific emphasis upon the frontier district and the effect of
constant raid and counter-raid. In Haldon's assessment, Leo was barely able to
22
Trcadgold,7.
1,2
out that his victories against the Arabs were largely as a result of the distraction
provided by the Abbasid Revolution. Haldon claims that after the revolution the
Abbasid regime came to accept the existence of the Byzantine Empire and thus
As may be seen from the preceding pages, the Isaurian emperors have
generally been treated in the light of their religious programs. This view has
often led to a negative assessment of the reigns of Leo and Constantine entirely.
I¡Vhile it is certainly the case that iconoclasm was an unpopular stand for the
emperors and did cause them difficulties, it does not necessarily follow that the
other policies pursued were also failures. In relation to the military situation in
the eighth century this is especially the case. Neither Leo nor Constantine can
realistically be faulted for the inability to conquer the Caliphate, or even for the
Ioss of Italy and Rome. The genius of the Isaurians lay in their ability to perceive
what was within the realm of the possible and to strive to accomplish it. It will
be demonstrated that Leo pursued a cohesive policy concerning the Arab Wars:
a policy which benefitted the Empire and one which the Empire was capable of
231.F.
Haldon, Byzantium in the Seventh Centurv. Cambridge University Press,
Cambridge, England, 1990, 84.
'n w,94.
73
Muslims that Asia Minor no longer was so accessible as it once had been. The
analysis which follows will attempt to delineate the military programs of first
Leo and then Constantine, as well as to determine the success or failure of these
programs.
14
m
BYZANTINE SOURCES
I.r order more fully to conprehend the relationship of the various sources
First Theophanes: the best known of the Byzantine chroniclers that deal
with the eighth century, he was a monk, born sometime betweenZ12-60.2s FIis
spatharios (a minor court official), and then shortly after the death of Leo fV
founded a monastery near Sigriane. There he lived in poor health until 815-816,
which saw the outbreak of a new bout of iconoclasm. Punished for his
2s
The biography of Theophanes presented here draws upon the more detailed one
provided by H. Turtledove, The Chronicle of Theophanes, University of Pennsylvania
Presr Philadelphia (7982), vü-ix.
15
other than the Chronographia. He wrote within the tradition of the chronicle.
Chroniclers generally wrote for a less educated audience than did the historians
of Byzantium, and consequently did not use the esoteric and archaic prose
period and did so in less detail, giving a year by year account of events.
Whereas Prokopios confined his record primarily to the events of the reign of
]ustinian and recorded these events in great detail, Malalas' chronicle begins
with the creation and continues to the death of Justinian I (14 November 565).
Prokopios thus: "Byzantine historians dealt with discreet churks of time, usually
a half century or less, which they treated with considerable detail".2ó TTìe
historian took it upon himself to address the questions raised by the text
concerning the events described, their causes and effects. Byzantine chroniclers
were rarely troubled by such matters, preferring instead to set down the
26 ibid, ix.
1.6
judgement.
hundred lines per yeaî, and often considerably less. It deals with the period
continuation of the Chronicle of George the Synkellos, another monk who began
his ch¡onicle with the time of Adam and had reached 284 A.D. by his own death
(c.810-11).2' Tlne date of completion for the Chronographia is not known with
absolute certainfy, however it must have been after 813, the terminus post quem of.
the work, and before 81-8,and the death of Theophanes.2s Up to the year 602,
602-813 none of the original writings which Theophanes may have consulted has
survived, indeed few sources from this critical period have survived the
dislocations of the seventh century, the iconoclastic purges of the eighth, and the
reaction of the iconophiles in the ninth.2e The period of interest for the present
27
ibíd,xi.
28
Turtledove believes that the Chronographia must have been written after 815, based
on Theophanes' statement on the current iconoclastic activity of the then reigning þut
unspecified) monarch, xi. This would coincide with the resurgence of iconoðhsmunder
læo V in 815.
2e
Turtledove, xv; who also notes that: "later Byzantine Chroniclers, even such eminent
men as George Kedrenos in the eleventh century, and John Zonaras in the twelfth, seemed
to have used Theophanes as their guide rather than the primary sources he himself
employed"; iùid, xi. He further mentions that Theophanes was known to the West as
early as the second half of the ninth century in a Latin translation; ibid, xvüi.
17
study is from the early years of the eighth century, which saw the rise of the
future Emperor Leo If[, to the death of his son Constantine V 1r.775.
This period of intense conflict over the use of icons in Christian worship
was also a crucial period in the expansion of the Islamic territories and in the
to the icons and the equally rigorous iconclasm of the Emperors in questiory one
important when deaiing with the final period of the Chronographia, from the
Theophanes' greatest source for these years is his own memory of the reign of
of a late eighth-century Syriac work which has not survived.3O Had this work
omitted details concerning the eastern border region, or whether his source was
incomplete.
The only other major Byzantine source for the eighth century is the
30Turtledove, xv.
18
exiled to Pontos in the 760s. Later Theodore was recalled but refused to deny his
iconophile sentiments and was again exiled, this time to Nicaea. Nikephorus
received an education that prepared him for official service, which he entered
and became an asekretis under the future Patriarch Tarasius, probably in the
Nikephorus continued in various posts within the civil service until the
death of Tarasius in 806. The Emperor Nikephorus (802-11), after consulting the
through the orders of monk, deacon, presbyter and bishop and was confirmed as
of Chuysopolis but soon after was forced to move to the monastery of St.
Theodore's. Michaeln @20-29) offered him the opportunity of return to the post
3t
The folllowing biography of Nikephorus is abstracted from Mango's introduction to
his translation of Nikephorus, Patriarch of Constantinople, Short Histor)¡, trans. C.
Mango; Dumbarton Oaks texts no.10; Dumbarton Oaks Research Library and Collection,
Washington,D.C. 1990; 1-16.
19
refused, continuing to lead the ljfe of a monk until his death in 828.
three of which have survived. Two are polemical pieces which, says Mango:
Both contain largely the same material as is found in the Breviarium, though
some minor items appear which are of little interest presently. Nikephorus also
composed a set of cfuonological tables, but his most historically significant work
The Breviarium treats the period from the accession of Phokas (602-10) to
who wrote within the tradition of the ch¡onicle, Nikephorus wrote a history and
aimed for a more sophisticated audience. Thus, he was bound by his use of this
32
These are the Third Antirheticus (PG 100,493 Dff;, and the unpublished Refutatio et
aversio. Mango,2.
33
Two redactions of this work are extant. The Vatican MS is the more complete of the
two, covering the period stated above. The London MS ends abruptly lr.713 with the
overthrow of Phillipicus (711-13). Both MSS are completely silent concerning the reign of
ConstantinelY (641-668). See Nikephorus, Short Historv. trans. Mango,5.
U'
34 g
1bld, .
20
internal evidence that can provide a date for the writing of the Breviarium.36
source for their treatment of the period under discussiory one which he claims: ".
the oddity that, even though Theophanes and Nikephorus drew upon the same
sollrce/ they were writing entirely independently of one another. Th"y were
service within the capitai at the same time. Mango speculates that both works
remained uncirculated for some time after their completion and that the
schoiarly circle of the Empire may not have been as close as modern researchers
3s
lbid-7; in tum citing P.J. Alexander, The Patriarch Nikephorus of Constantinople,
Oxford (1958), 157-8. Theophylact of Simokatta wrote a history which continued that of
Prokopios and ended with the murder of Maurice and the usulpation of Phokas n 602.
see, Theophylacti Simokattae Historiae. ed. C. de Boor (Leipzig,1ß97).
36
Mango tentatively suggests a date sometime in the 7B0s; p.1,2.
37
ibid,9; who adds, " From 668 to the end, the Breviarium and Theophanes run in
parellel channels and are quite clearly derived from the same source. To be more precise,
there is very little in Nikephorus that is not also in Theophanes, whereas the latter
includes a considerable body of other material, some of it Near Eastern in origin . . . The
source was certainly a chronicle composed in the capital and it appeats to have been
favorable to læo III.", 15.
"¡bid 12.
21,
concerning the iconoclast Emperors. With the exception of a few minor sources
such as the Chronography of Gregory abu Faraj (Bar Hebraeus, d.1286) and the
work in eighth and ninth century Byzantium, the dearth of other Byzantine
sources forces one to go farther afield and look beyond the boundaries of the
Empire.
22
IV
SOURCES - MUSLIM
with Theophanes. Flowever, several works that are now lost have been
necessary.
elements. The earliest historical writings consist of what Somogyi calls: "the
paganism (al-akhbar) on the one hand, and the geneological studies (al-ansab)
on the other".3e With the conversion of the Arabs to Islam the already ttriving
written traditions of 'al-akhbar' and 'al-ansab'were adopted and modified to
become one pursuit with two foci - traditions conceming the life of the Prophet
From this interest in the rise of the Prophet and of Islam evolved the
3e
Somogyi ,1. de, "The Development of Arabic Historiography", Journal of Semitic
Studies. IIl, 1958, 37 4.
40
ibiù gr4.
23
practise by an isnad. In fact, hadith are made up of an isnad and a matur. The
isnad is a chain of authorities which originally reached back to some source that
was in the presence of the Prophet, and who ostensibly is the one speaking in the
matn. Williams gives an excellent example and decription of the hadith. isnad
and matn:
Early Muslim chroniclers used the same hadith format of isnad and matn
a1
Læwis, 8., ed. Islam: From the Prophet Mohammed to the Capture of Constantinople,
Oxford University Press: Oxford, 1974, Volume 2,xvä.
4 Al-Tabari, The Earl)¡ Abbasid Caliphate, trans.
J.A. Williams; Cambridge University
Press, Cambridge (1988), vol.1 of 2,xvi;
These hadith are the basis of Muslim jurisprudence, just as the Quran is the basis
of Muslim law. An individual who interprets the hadith is called a faqih, and may inspire
a foìLowing who hold to the same intelpretation. ln this way, Muslim schools of law
formed. Tabari inspired one such school; his enemies, the Hanbalites, were the followers
of another, founded by Ahmad ibn Hanbal.
24
when they wrote history.4 Thus, a typical entry in a Muslim record purports to
chroniclers who made use of the hadith style generally included any hadith
which had at least a plausible isnad. Thus it is not uncorunon to find different
accounts of the same subject in the work of one chronicler; because the clrronicler
relate to the year's events. Often one or more of the hadith concerns the same
same way that modern footnoting provides a method for the corroboration of
information. At the end of each yearly entry, the chronicler would often include
whichhave no isnad, and are often the remarks of the chronicler himself.
n' Lewis,
I€tae Volume 2,xx, suggests that the adoption of hadith by historians and
ctuoniciers was encouraged by the early Ummayad rulers, who wished to increase the
authority of their position through association with the Prophet.
25
continuous narrative without any use of the isnad and matn, or of individual
hadith. The writer took the various hadith with which he was familiar and
transformed them into a single homogeneous account which best represented the
information in his possession. Neither Baladhuri (d.892) nor Tabarí (d.923) were
of this traditiory but were almost certainly familiar with it. Masudi (d.956) had
already adopted this style. Thus, by the beginrring of the fourth century of
counterparts.
somewhat differently. In these texts the author organized his subject district by
each region. Within the account of the region events generally were dealt with
chronoiogically, but dates were provided less often and the entries were not
arranged as annals.
would today be called human geography and often recorded details of the
cultures, local history and various anecdotal stories of a region rather than
his interest in the history of the Caucasus provides theByzantine historian with
26
(dating to the tenth century A.D.) is also of use in this regard. With an
and a native of Baghdad. His grandfather was a court functionary to the Caliph
which is now lost. He also wrote two other works, one of which survives only in
adopts the style of the continuous narrative which became popular in the
following century. It records the Muslim conquests of the seventh, eighth and
chronological format.
44
Al-Baladhuri, The Origins of the Islamic State. Columbia University Press, New York,
1.9L6,p.7;
For the sake of clarity, I will attempt to render the Arabic or Persian names as
clearly as possible into English, without the inflectionary notations and with as few
patronymics as is possible. I will also attempt to regularize the references to various
individuals. As an example: The Caliph al-Mansur (754-75) is known to Theophanes as
Abd Allah; Tabari refers to him only rarely as al-Mansur (the name he chose when he
became caliph), instead using 'Abu Ja'far' (and very rarely 'Abd Allah'). It should also be
noted that'b.'is the recognized abbreviation of ib¡, (son of).
27
Abu ]a'far Muhammed lbn ]arrir al-Tabari was born of a Persian family of
had an interest in the hadith from an early age and moved to the town of Rayy to
further his study. At age 30, Tabari settled in Baghdad to collect more hadith.
From there he moved to Basra, then on to Wasit and then to Kufa. While in Kufa
continue his education in hadith, and to investigate the teachings of the various
law schools. Later Tabari moved to Syria and then to Egypt in867, where he
Hanbal (another man famous for his knowledge of hadith; d. 855). This
escalated into a large-scale mob action on the part of the Hanbalites, which it
took thousands of troops to quell. Tabari's life was made extremely difficult by
the Hanbalites, who slandered him and kept scholars from seeking him out as a
source of hadith. He went into hiding, and it was during this enforced seclusion
that Tabari wrote. Eventually the furor died down and he was able to return to
public life. His opinion was much sought as a jurist and hadithist. He died in
Baghdad n923.
Kings') records the period from creation to his own time and was written in the
tradition of the hadith which affected virtually every aspect of Tabari's hÍe.as
text which describes the state of the Caliphate and its neighbors in Yaqubi's own
time (c. 900 A.D.).46 FIis historical work, Al-Tarikh, has much to add to the data
nowhere else.aT
Various other Muslim sources are of more limited utility for the present
Persian extraction dating to the tenth century. Minorsky suggests that it is the
45
Published as Tabari, Annales, ed. M. J. de Goeje, et. al. læiden, 1879-190L, L3 vols.
The English translation of the text is presented by: State University of New York Press in
38 Volumes with various translators and editors. Henceforth, these volumes will be cited
by translator and page number. Another translation of the period 754-809 has been
provided by l. e. Williams: al-Tabari, The Earl)¡ Abbasid Caliphate. trans. J. A. Williams.
Cambridge University Press: Cambridge,7988. 2 vols.
a6
Al-Buldan, ed. M.J. de Goeje, læiden, 1892. Translated as l.es Pa)¡s, G. Wiet, Paús, L937.
47
Al-Tarikh, ed. M.T. Houtsma, Leiden, 1883. Al-Tarikh is available ouside of Arabic
only through Brooks in exce{pted portions of the text. All of Brooks' references to Yaqubi
(or, Ibn Wadhih, as Brooks calls him) are to Al-Tarikh. E.W. Brooks, "The Arabs in Asia
Minor from Arabic Sources, 641,-750",IH918 (1898) 782-208.
aB
Minorsky dates the compilation of the text to 372 A.H. (9S2 A.D.), Hudud al-
Alam, xiv.
29
manuscript, the text survives in only partial form. The extant portion of the
Kkritab al-Uvun begins in705 A.D. and is, according to Brooks, the sole
From the above analysis, it is apparent that the addition of the A¡abic
ae
Brooks, "Arabs in Asia Minor, 641-750," 182.
30
LEO Itr
end a rapid and violent succession of emperors which saw six different reigns
Officially the Empire had been at peace with the Caliphate since the
Circumstances forced the Caliph Muawiya to sue for peace in order to allow
the armies of Islam to recover from the disaster. Muawiya's successor, Yazid
(680-83), saw no reason to alter matters since much of his might was involved
in putting down a revolt in the East. Upon Yazid's death, civil war erupted
until Marwan established supremacy, passing the Caliphate on to his son, Abd
Abd al-Malik promptly renewed the treaty with Byzantium under the
original terms; he then set about healing the internal divisions of the Caliphate,
suppressing a series of annual rebellions until 691-2. It was at this time that
relations grew strained between the two realms. Abd al-Malik attempted to
continue amicable relations with the Emperor ]ustinian II, the successor of
the new Emperor, but whatever the ïeason, the Empire and the Caliphate met
31
on the battlefield of Sebastopolis in 692-9. The outcome was a defeat for the
imperial clientship and transfer his allegiance to Abd al-Malik in the following
year (692-3). Theophanes concludes his entry for that year by saying: From
"
beginrLing ín 694-5. After ]ustinian's fall from power n 695-6, matters rapidly
grew worse. Sergios, the Patrician of Lazika,revolted against the Empire and
placed himself under Arab rule. In 697-8, after brief resistance, the Muslims
conquered Byzantine Africa. That same year, Apsimar, the Drungarios of the
seized the Imperial tluone from the ineffectual successor of ]ustinian, Leontios.
against the raids of the Arabs. Though he made no move toward the recovery
duties competently, even managing to take the offensive in 700-1', and routing
the invad.erc in70t-2 and 704-5. Flowever, he was not able to stem the tide of
raiders. Taranton (Tarandos) was captured and settled 1n701-2- Tn702-3, aI-
5o Theopharres,64.
32
tsaladhuri notes: "The Muslims had never lived in this town before."sl Lr
addition to these inroads, some time between 690 and 705 Kaisariyyah and
the Empire. The region of Fourth Armenia went over to Arab rule in 7024.æ
When the Armenians attempted to return to the Byzantine fold the next year,
they were unsuccessfut and brutally punished by their Arab masters. Thus,
with the return to power of ]ustinian II, the Empire had lost the Armenians
He went on, either in the Same year or the next, to conquer Tuwanah (Tyana),
while Hisham is said to have taken the unidentified fortresses of Buluk, al-
s1 Baladhuri,255.
Akhram, Bulus and Qumqum.ss Baladhuri aiso records the fall of Tuwanah
plrrposes the precise dates of the events are not as important as their sequence.
On this Tabari is clear, The conquest of Tuwanah and the variously named
citadels paved the way for a major invasion through Cilicia and deep into the
hinterland of the Empire itself in the period immediately after the actions
outlined above.s6
(Herakleia) and continued into Isauria where Tabari states that he conquered
the fortress of Suriyah (the citadel of Isaura or an unnamed fort within Isauria
55 Tabari, Hinds, 134; Hinds is obviously correct in his belief that there are
chronological problems here. Tabari lists the conquest of Tuwanah under both 705-6
and706-1. Unãer the latter entry Tabari gives the captured fortresses as Qusta¡tin,
al-Ghazala and al-Akhram; Tuwanah/Tyana is mentioned as the conquered city in
both entries, but Buluk, Bulus and Qumqum are given only in the first. Qustantin is
Constantine, north-west of Edessa. The rest defy identification and one or more may
be altemate names for another in the grouP - Bulus and Buluk, for example'
Qumqum may be the Byzantine fortréss of Komacha-ani/Kamachon (generally known
to the Arab sources as Kamkh), but the location does not fit the described events'
st For simplicity, Tabari's chronology and sequence of events will be accepted with
reservation. Tabari's record is the most complete in addition to being the most
confused. Interestingly enough, a similar chronological inconsistency may be found
when comparing the entries fór the siege of Constantinople in 716-1'8' Perhaps the
sources wfrictr Theophanes, Tabari and the others had to deal with were so numerous
and contradictory as to render dates essentially meaningless.
34
Masiamah, who would have had to pass Podandos to get to Herakleia and was
around the city while another force continued on to assault Dorylaion. Al-
Abbas met and routed another Byzantine army and took the city itself.
It is also possible that Al-Abbas set off north from Podandos while
Kamouliana, but it is possible that Tabari is misinformed on this point; also the
heading east then south to avoid the difficutty of having to provision his
troops off the same settlements which were dislocated or raided on the
5?
Tabari, Hinds, 1,46; lt is not recorded whether Amorion fell.
ss
Al-Yaqubi provides no significant details for this year. But he records the use of
the two-prõnged attack on several later occasions. For 723-4 A.D.: "Abd-al-Rachman,
the son óf Sot,orno.t, the Khalbi, made a raid on the south in the summer; and
Uthman, the son of Chayyan, the Murri, made a raid upon the north in the summer."
(Brooks, "Arabs in Asia Minor, 641-750," 197-8. Similar entries appear for the years
726-7 artd729-30.
35
westward march.
In the decade which followed, the forces of Islam did not falter in their
Constantinople (776-8) only one year did not see a significant Arab raid upon
inhabitants who, says Tabari: " . . . migrated to the inner part of the Byzantine
Masah and three forts near Melitene, and Marwan b. al-Walid reached
Hiraqlah fell for the second time a year later. The incessant raiding continued
may have been premature for the late 680s, but it is evident that the Caucasus
situation deteriorated greatly over the last years of the seventh century and the
early ones of the eighth. By the second decade of the eighth century the Arabs
did hold many of the 'cities of the heights'; so too did they control the access
to the Caucasus passes. Thus the Muslim raiders were able to move virtually
at will within imperial territory. By 71.6, the forces of Islam were operating
well within the realm of the possible in their attempt to take Constantinople'
The sources which report on this momentous event are many, but the
the most confusing) source is that of Theophanes. Nikephorus has little to add
that is not in Theophanes' wotk. Nikephorus has either derived his own
case because Nikephorus does not include the difficult passages found in the
Nikephorus may also, directly or indirectly, have had access) for the events of
the siege is evident. Turtledove notes that, in describing the siege, Theophanes
refers to Leo III as: " . . . the pious Emperor . . .".64 This is, as Turtledove
correctly points out, the sort of error that Theophanes could only have made
had he been copying rather carelessly from what was probably a text of a date
Theophanes' account of the siege provides a good basis for comparison with
other sources which may be used to shed more light on the difficu-lties in the
Chronosraphia.
.+
Constantinople, yet Maslamah did not succeed in doing so until the following
year. While Constantinople must have been his objective, and Theophanes
mentions that preparations had been made for that purpose, it becomes
obvious that Leo, then strategos of the Anatolic theme, was instrumental in
wrote to him to discuss terms.6ó The Arabs taid siege to Amorion and
attempted to negotiate with Leo, but Leo dragged out the talks and escaped a
trap set for him by Suleiman. Leo was so successful in his efforts that the
places elsewhere.6T
and moved the non-combatants out. The main Muslim force, under
Maslamah, appeared at Amorion just as Leo left for Pisidia. Leo continued the
negotiations for an alliance with Maslamah, at the same time letting Maslamah
know that Amorion had been reinJorced and would not fall easily. Evenfually,
Leo could stall no longer but Maslamah was forced to move on in order to
reprovision his force. He took his army to Akroinon and then captured
Pergamon. But by then it was the end of the raiding season and Maslamah
took his troops back beyond the Imperial border. Leo had effectively diverted
where he captured the Emperor Theodosius'son as well as the: " ' . . entite
6TTheophanes,
83.
ut
The significance of the delay should not be underestimated. Theophanes tells us
that Artemios sent emissaries to spy upon the Arabs'preparations and thus was
aware of the coming invasion befõre the death of Al-Watid (late February,71'5,
according to Tabaril Hinds, 21.8;). k is also recorded that Artemios used this
knowledle to improve the defences of the lmperial City by expanding^the fleet,
rebuildin"g wail sãcHons and constructing siege engines; Theophane-s,80' It may well
be that Ló bought the Emperor crucial time io complete the refortifications. As it
turned out, Mas-iamah did not begrn the actual siege of Constantinople until 15
August 777,the year after his encounter with Leo at Amorion'
39
imperial retinue and the palace's leading fígures."óe Theodosius was convinced
by these events to relinquish the tfuone. Leo's route to imperial power was
clear.
began its major campaign with renewed vigour. He again attempted to get
Maslamah to confer with him, but the A¡ab general was rightly suspicious.
Suleiman with the rest of the expedition. The siege began on l-5 August777
and lasted a fullyear. The landward sides of the city were blockaded by a
ditch and. rampart. Suleiman arrived on the first of September with L800 ships
fult of supplies and troops. This huge armada was split to find harbour in the
various ports along the Bosporus near Constantinople where they could
effectively seal the city from maritime contact. But unfavourable winds for the
Muslim fleet allowed Leo to send out fire-ships against the vulnerable invaders
and many of the Arab vessels were burned. Crucial provisions and men were
lost and the besiegers could not afford to lose more ships in an attempt to
Due to the damage done to the naval force and the loss of supplies
incurred thereby, the Arabs faced a winter of hardship, famine and disease.
6e
Theophanes, 34. Atthough Theophanes does not say so, it is to be expected that
there waian army with the Emperorls heir that [.eo either defeated or subomed. It is
unlikely that such a collection ôf personages would have been outside Constantinople
on the main military route while ã major invasion and an untamed strategos roamed
the hinterland.
40
These conditions combined with uncorïunonty cold weather to kill many of the
besiegers as the months wore on. Meanwhile since the A¡ab ships could not
seal off the Cify, provisions could be brought in, and although it would not
have been an easy time for Constantinopolitans, they had only to look beyond
from Egypt. The ships were guarded closely and harboured secretly to avoid
a disaster similar to that of the previous year. But the Egyptians of this new
armada rebelled and gave the Emperor enough information that he was able
once again to fire the Arab ships and carry away considerable plunder.
A land army had also been sent to the aid of Maslamah, but it was
disastrous venture the Bulgars attacked the besiegers and slaughtered a great
number of them.7l The remainder withdrew and suffered further losses due to
72
storms and other disasters until few refumed to their homes.
Theophanes' tale of unending woe for the forces of the infidel is almost
70
Theophanes, 80, relates that Artemios commanded that all who would stay
within the City walls during the siege should have three years of provisions on hand.
Thus there is little doubt thát those inside the City were better off than the invaders.
justice and the intervention of the Virgin to save the City, it is apparent that
the version of events that Theophanes has preserved is, to an unknown but
equatly troublesome. Leo foois two able Arab generals with rather transparent
cred-ible as that of Theophanes. Leo ínitiated negotiations uPon the entry of the
Islamic force into the Empire. His intent was simply to bribe the army to
offer of one dinar per soldier was refused and the talks end with no further
citizens into surrender, he should destroy the provender of the besieging army.
This would serve to demonstrate the intent of the army to storm the walls
rather than settle down to a lengthy siege. Maslamah followed this highly
dubious council and so began the famine mentioned in all sources concerning
the siege.
t'As an example, during the negotiations with Suleiman, Leo reportedly came into
the camp of the enemy with only an escort of three hundred men. He was let free
but knew that the Caliph wanted to capture him. After sending ten times the number
of Leo's own force to sur¡ound him, the Chronograptúa has Leo escape by feigning
that he is going hunting, and instead shifting his camp. When accosted by his
pursuers he is allowed simply to go on his way. Theophanes,S2-3'
42
An alternate version of the event from the same soulce has Leo offering
to deliver the Empire to Suleiman and then tricking Maslamah into allowing
the besieged citizens an opportunity to gather food outside the city proper' In
tl'Lis way, Leo suggested to Maslamah: " . . . the peopte would believe that
Leo's word and Maslamah's were one and that they were safe from being
captured."Ta Once the citizens were provisioned, Leo: ". . . began to act in a
shame even a woman."7s This in some way led to the famine and the dire
which satisfies the suspicions of the modern reader. The ruses recorded in
either source could hardly have succeeded as they are Presented;yet, it would
historiography, that Maslamah was clearly duped by the future Emperor, Leo
m.76
7a
Tabarí, Powers, 41.
t5
ibid.
tu
As Gero points out, The anonymous Arabic Khitab al Uyun. an eleventh
century Spanish text, probably based upon much earlier eastern sources, also
pt"r"óejthe theme of Leo's deception of Maslamah, as well as his utter scom for the
Mushm commander; Gero, Byzantine Iconoclasm in the Time of Leo IIL 33n.
Also worthy of note ii Gero's mention of the favourable treatment of Leo in
the Armenian source material; material which would have been outside the iconophile
revisionisrn of the post-iconoclastic period, Gero, LeolIL37.
43
that it did not draw upon either as a source. Lr Gregory's work, Leo promised
With such backing Leo felt powerful enough to attack the Emperor Theodosius
at Amorion and take him prisoner. From there Leo went to Constantinople
back the Arabs. And they were very pleased with him and they made him
king over them."z Maslamah then appeared before the city, but was soon in
dire straits, caught between the forces of the city before him and Bulgar
mercenaries. Famine added to the host of Arab troubles and when the army
Gregory's version is in many ways the most satisfying of all the sources.
But perhaps the most sensible conclusion is in some amalgam of all the
information available. In any event, the exact details of the situation are not of
Abu Faraj: that Leo in some way outmanoeuvred and delayed the superior
In the events of the siege, regardless of how the details are interpreted,
7'
Gregory Abu Faraj, 108.
44
Leo demonstrated a shrewd ability to combine the use of martial power with
force was precisely timed to take advantage of weaknesses within the enemy
camp - weaknesses often created by Leo himself. This precision also extended
to the use of mercenaries and allies. Leo was well aware of the benefits of
having an outside agency fight his battles for him. As Emperor, Leo III
bear the brunt of enemy attack. Leo was not a latter day Herakleios and his
alliances of mutual support, of subtle influence exerted upon peoples who had
The reign of Leo III did signal the beginning of a new military
relationship with the Caliphate and without this change the Emperor's policy
tt Ninth
J.F. Haldon, H. Kennedy, "The Arab-Byzantine Frontier in the Eighth and
Cenfuries:MiIitaryorganizaiionandSocietyintheBorderlands,.'@
Vizantinoloskog lnstituta. 19 (1980) 82.
45
After the siege the situation was changed. Through a combination of carefully
appiied military pressure and the nurturing of alliances with peoples with a
common cause, Leo was able to fortify the Empire and Pave the way for its
eventual rejuvenation.
the Faithful. With his direction in 7L8 that the siege of Constantinople be
called. off Umar adopted a new and more cautious policy concerning the
Byzantine situation. But his change in attitude was not due solely to the status
of the siege itself. Rather it was his concem over repeated attacks made along
the frontier by the Byzantines. With this in mind,Ibn al-Athir relates that
unoccupied, through fear of injury to the Moslems from the enemy'"7e lJmar
seems to have felt that the border-towns were too exposed to attack, for he
7e
Ibn al-Arhir, in Brooks, lgg 18 (1898) '1.97. Baladhuri adds that Umar had
Turandah destroyed; p.290.
46
Apparently Leo launched an attack into the region shortly after the siege
t Baladhuri,255.
tt They could have been targets of a single campaign, either land or sea-based. It
seems more likely that a two pronged attack was made - with one force moving
through the Cilician Gates and the other through the pass of Adata.
82
Baladhuri,204;he also notes that Yazid II (720-24) increased the size of the
garrison. Yaqubi confirms the attack but states that it occurred while: " . ' ' the
Moslems werã smitten by scarcity, and hunger, and cold. . ." befote the gates of
Constantinople; Brooks,IÞ 18 (1898) 195.
47
Caliphate. Umar thought the situation desperate enough to pay tribute to the
continued attack. That first Umar and then Yazíd saw the necessity of
military inadequacy.
Under yazíd,U, (720-4) the Caliphate was able to recover somewhat from
once more and Muslim victories again began to occur. After no Summer
raids
resulting in the defeat of. aByzantine field force in Fourth Armenia and the
taking of a city.æ For the year following, Yaqubi mentions that al-Abbas led a
raid into imperial territory only to have his command destroyed piecemeal as
æ
Ibn al-Athir has Dalisa which Brooks suggests might be Dalisandos, THS' 18
(1898) 197. For727-2,Tabari records the samJcommander taking Raslah, which
'!'67. Powers also notes that
Èo*e6 identifies tentatively as Larissa; Tabari, Powers,
another Arab chronicler, al-'Azd.i, gives the city as Awasa, 167n. The l{ritab al-Uyun
records the fall of Muwasa for the-year 7?3-4, Brooks, THS,18,(1898) 198. Since none
of the authorities mentions the fall of more than one city, it is not unlikely that Dalisa'
Raslah, Awasa and Muwasa all refer to the same place' It is also possible, given
the
rather convoluted chronology of these entries, thãt Dalisa and Raslah refer to the
events of one year while Añása and Muwasa correspond to the actions of the next'
&
Yaqubi, Brooks, IHg18 (1.898) 197.
48
InTZg- ,a Muslim army was again caught after having separated into
I+ramkh and Kuniya.s6 Theophanes' entry for the year contradicts those of the
Muslim authorities. The newly installed Caliph Hisham is said to have led an
attack but apparently: " . . . he withdrew after squandering many of his men'"87
tt Tabari, Powers,192.
is
Obviously the different accounts present some difficulfy' Theophanes
relate an
probably referring to the same raid as Taban,but Yaqubi appears to
attack was launched by the Muslims with one army heading north
from
have fallen and one of the two prongs dispersed into looting parties only
to be
Whether or not Ikonion was attacked, it is clear from the following entuy
that the Arabs derived no strategic advantage from the year's activities which
carried over to the next season. Fo¡t 724-5,Tabari records a raid- of little
significance. Yaqubi has more detail, stating that the Muslims razed some croPs
and villages in retaliation for the Byzantine destruction of enemy pasture lands'e0
This year's events suggest that the situation was in something of a stalemate with
each side carrying on raids of a very local and limited nature' Yaqubi records
a
and extensive, and at first glance aPPeaf to herald a return to the devastating
8e
If this is indeed how events occurred then it makes the most sense to assume that the
Ciljcian arm of the campaign proceeded to the environs of Ikonion and, perhaps
blockading the city, setãbõutþiltag*g only to be caught by the Byzantine response'
e0
Tabari, Blankinship, 8; Yaqubi, Brooks,lFIg 18 (1898) 198'
50
Theoohanes and the Khritab al-Uvun all tell of the fall of Caesarea to Maslamah,
L
this attack was roughly coincidental with two other assaults. Flisham's son,
Muawiya, reported,ly took his command: " . . . here and there and withdrew."e2
-
el
Tabari, Blankinship,2T,andTheophanes,g6,place the capture n726-7. Khitab
al-U}¡un. rooks,IFIS, 18 (1898) 198, has 725-6. A major attack upon Caesarea does not
lend support to the likelihood of the fall of Ikonion two years earlier; rather it argues
against it. If Ikonion had fallen it would have provided an advantageous ståging point for
further penetration of Byzantine temitory. Thus Muslim raiding in subsequent seasons
should have concentrated in the region of Ikonion itself.
Incidentall/, Abu Faraj records not Caesarea but Neo-Caesarea, however he is
alone in this regard, 109. The Khitab al-Uj¡un locates Kaisari]¡ya between Malatyah
and Khamkh, Brooks, THS 18 (1893) 198. While this locale does not exactly describe
either Caesarea or Neo-Caesarea it is more applicable to the latter.
e?
Theoohanes. g6.
lr'
51
Nikeohorus' Breviarium also mentions the Nikaia attack (under the year 727-8)
but does not give so negative an impression as Theophanes:
In these two passages it would appear that Theophanes and Nikephorus are
drawing from the same source. The details and structure are alike, but
Theophanes has obviously fitled out his account with some polemical prose
referring to Leo's issue of the edict against icons (726 A.D.).es Otherwise there is
little to distinguish between the two records other than the somewhat suspect
Nikaians.
e3
Theophanes, gS. Apparently Theophanes is relying upon more than one source
for this et^,t.y, and copying rathei sloppily from them. First Theophanesrelates
Muawiya's uneventfuÍ raid, then he give details which make it obvious that Muawiya's
raid was anything but uneventful.
e4
Nikephorus, Mango , c.61,p.131; Mango provides information which may
corroboraË a less pessilristic version of theliege of Nikaia. An inscriptiol {a1ing to the
time of læo III comme-orates the failure of theArab attack, p.212., citing A.M. Shnieder
and W. Kamapp, Die Stadtmanor von Iznik (Nicaea), (Berlin, 1948) 49,no'29 and p1.50. _
It should also be noted that the Muslim authorities either had no knowledge of the
attack on Nikaia or did not think itworthy of mention; see Nikephorus, Breviarium'
Manga,212.
e5
M.V. Anastos'arguments in Cambridge Medieval History, 4,part 1 (second
See
edition, Cambridge, England, 1966).
52
Gregory Abu Faraj provides information which also suggests that this
year saw a major campaign. Abu Faraj records the siege of Nikaia [albeit tmder
the year 73tl as being immediately after the capture of Gangra in northern
Cappadokia. Yaqubi con-firms the fall of Klrangara in the year 727-8, which
and possibly Khamkh moving down the Byzantine military road into the
reconstruction represented actual events then this season was without a doubt
the most devastating for the Byzantines since the accession of Leo III.
disastrous than were the Citcian campaigns of the years prior to the siege of
believed),and Caesarea, Kuniya and Gangra were sacked but not held.e8 The
only actual loss was the citadel of l(hamkh/Khamacha-ani and this fortress
e6
Abu Faraj, 110. Yaqubi, Brooks,Ire, 18 (1898) 199.
eT
It is not unreasonable to assume that Khamkh was at this time in the hands of
the Caliphate. lnterestirgly, Neo-Caesarea is also directly on this route and seems to
fit the probable campaign scheme well. However it is as likely that Caesarea was taken
by a different force under Maslamah while the main force travelled down the northern
rnilitary road under Muawiya.
e8
\Alhatever the identity of Kuniya there is nothing to suggest that it remained in
Muslim hands.
53
wouid remain hotiy contested by the Empire and Caliphate for some time to
come.
must be noted that both Theophanes and Nikephorus stress the awesome size of
the Muslim force involved in the campaign.Ð That the Caliphate could raise an
army several times greater than could the Empire is to be expected' Yet the
Muslims lost men in the field time and again, thus the Byzantine response was
more than simply to shelter behind fortifications. Leo III or the strategos on the
scene was able to organise a sort of guerilla resístance even to such a massive
assault.
In the following season (728-9) the attacks apPear to have been more
en
It is quite easy to imagine an expedition of the scope of the siege of Constantinople
nTIT-8. Éor that army the;crual numbers are unknown, but the naval contingent alone
is said to have involveã a total of 2460 ships - and this comprised only the supplies and
reinforcements for the main army which travelled overland and was itself also reinforced
by that route (Theophanes, SS-89). If the estimate of 1.00,000 men is taken as an accufate
aésessment of the army's strength, the army could well have outnumbered the entirety of
the Byzantine military. While the actual numbers need not be taken literally, it is
apparent that the Byiantine writers were impressgd by the size of the force involved in
tt",ö Nituiu campaign. Obviously, Theophanãs and Nikephorus (or their colrunon source)
intended that tñe size of the inváding force be perceived as overwhelming' Treadgold
notes that the largest Byzantine field force recorded in the eighth cenrury was 20,000 men,
w. State Finances in the Eighth and Ninth
B)'za¡tine state
W. Treadgold, By-za¡tine Nmtn cenrury,
Century, tsast European
East EuroPei
Monographs, Columbia University Press, New York, 1982,92'
- j.
Hatdon, on the basis of tire testimony of lbn Khurradabih's geography, Kitab
al-Masalik wa'l Mamalik. estimates the Byzantine Tagmata (the regular ailny stationed
at ttre iapital) to number no more than 8,Ó00 in the eighth century; "K-udamah Ibn
Djafar ur1¿ tt Garrison of Constantinople," Byzantion, 48 (1978) 80. Yaqubi states
"
that for his own time the thematic army contained only 40,000 cavaky, Les Pays. ch'323'
54
[mited in scope.100 Land and sea raids resulted in the capture of a fortress, but
saw
the Muslims again lost men to aByzantine counter-attack.l0l The next year
the falt of samaluh to Muawiya, and more sea raids.1o2 Another series of
raids in
72g-30was uneventful, though Tabari reports one unit: " . . ' even reaching
burned'le
Qaysariyyah."lo3 In 730-1 lharshanah was captured and Farandiyyah
In either this year or the next, the Byzantines again inflicted a significant
,defeat upon the invaders. Tabari's account is the most informative: "Abd al-
wahhab b. Bukhr went out campaigning with al-Battal in the year II3I73L-21,
but al-Battal's forces were defeated and fell back."lOs In addition to this defeat,
too
Haldon and Kennedy, "Arab-Byzantine Ffontier," 115, who Poin! oYt-the
See
enoÍnous expense involved in mounting an expedition of such a scale as the Nikaia
campaign.
101
Tabari, Blankinship, 33; the fortress called Tibah by Tabari,maybe Thebasa, 15
Miies northwest of Herikleia. For 728-9, the ICritab al-Úyun records the same details
of the raid but the name of the fort is lost to alacuna. Yaqubi refers only to the fall of
(1898)
Gangra, which is in all likelihood a part of the push to Nikaia, Brooks, THS 18
f 99. iheophanes records the city aJ Ateo..ts, which is unidentified but may be
synon)¡rnous with Tibah / Thebasa, Theophanes, 98.
toz
Tabari, Blankinship,45. Blankinship places Samaluh near Tarsus and Al-Massissa,
but he equates it to Semålous on the nortñem military road between Kharsianon and
Dorylaeoì. Obviously they cannot be the same. If Semalous is not meant, it would mean
that the Byzantines still helct a town in Muslim controlled Cilicia.
to3 Blankinship,64. Apparently it was still an
Caesarea in Cappadokia, Tabari,
accomplishment to vãnt.rre even into eastem Cappadokia at this time'
10a
Theophanes gives the commander as Maslamah, but Maslamah was at waf in
the Transcaucasus according to Tabari, Blankinship, T0. Kharshanah is Kharsia¡on,
Farandiyyah is unidentified.
10s
Tabari, Blankinship, g5. The Khitab al-U)¡un dates the event to the previous
year, Brooks, JHS 18 (1898) 200.
55
territory with his force this year, but remained instead at Marash.10ó It would
appear that the Byzantine presence in the border region at this time was strong
enough to defeat one raiding force and intimidate another into inaction.toT
process. Muawiya is said to have taken the outer city of Akroinon itself-
Another raid that year reached as far as Caesarea but is not said to have
1 08
accomplished anything.
For the next four years little is recorded by either the Byzantine or the
Muslim authorities. A campaign was made each year but only Theophanes
provides any more detail. For 73F-STheophanes says merely that: "Muawiya
106
Tabari, Blankinship, 95. Yaqubi records the same information to 730-7, stating
that Muawiya: " . . . did not succeeã in entering their [the Romans'] territory." Brooks,
THS 18 (1898) 200.
107If the altemate chronology of Yaqubi and the I(hitab al-Ulzun is favoured, then
Muawiya was able to pin the Èyzantine defensive force without exposing his command to
risk. Hâd Muawiya's ãr*y not been there the Byzantines might have been able to relieve
the siege of Karsianon after defeating al-Battal.
t08
Tabari, Blankinship,gT. There is no reason to susPect that this stategos Constantine
was the future Emperor-Constantine V (71,8-nÐ. Theophanes confirms a raid into
Paphlagonia, LOL. iaqubi places the defeat of strategos ConstantineinTSS-4, as does the
tGìtab ãl-Uvun. Brookì,IHS 18 (1898) 200. The Khitab al-Uyun does not mention a
Constantine b.tt .ecords events as disasffous for the Byzantines: " . ' . The Romans were
routed and the Moslems fell upon them and made great slaughter, and took many
captives and took possession of *te camp and made sport of their property'" Brooks,lH$
fg (f ggg) 20L. The }Gitab at-U)¡un is suspect, however, because it has merged this conflict
with the battle at Akroinon n73940.
Concerning Caesarea, Tabari may well be repeating his entry for the previous
year.
56
devastated Asia."læ Other than the occurrence of raids, no source has any
more
,detail for the period 735-6 to 737-8. For the last year, Theophanes states that
another source is for the year 798-9,when Maslamah raided the Empire once
the Empire.tl2 This time, however, events did not favor the armies of the
commanders. The other forces were successful in their raids but are not
recorded as having accomplished more than 1ooting.113 The surviving 6800 of Ai-
Battal's cavalry escaped to Sulieman and then to Syria. While the destruction of
one-sixth of the Muslim army may not seem crippting, it is unlikely that all the
looting that Sulieman perpetrated would compensate for the loss of equipment
and investment represented by the slaughtered thousands. Indeed, the next year
Yaqubi is the only authority which records any action in Byzantine lands, and it
Thus the campaigns of Leo III's reign, while not a story of uninterrupted
imperial success, demonstrate that the forces of Islam were no longer able to raid
With the exception of the offensive launched immediately after the failure
aggression went unrecorded, there is no reason to believe that any such events
occurred after Umar tr and Yazidtr took measures to refortify the border. From
this it may be concluded that, by LJmar's death (20 Februaty,720), the Muslim
defences had recovered and sufficient men were once again under arms to deter
Byzantine attack.
After the accession of lJmar's successor, Yazid II, the Byzantine strategy
changed. No ionger choosing to raid outside their own borders, ii:rrperial forces
entering Byzantine territory. The standard pioy used was of hitting various
detachments after the raiding force had dispersed to loot and/or forage'
A tenth
The actions of Leo's armies adhere very closely to the principles set down in the
Manual of Nikephorus II's time. Clearly Leo was able to assemble a competent
but not overly large force in time to attack the raiding detachments of the Arabs
as they attempted to loot the Byzantine hinterland. On only rare occasions was
Leo abie to accomplish the destruction of a large contingent of the enemy army
against a major Arab incursion Leo would have had to employ not only the
The assembly of sufficient forces to engage in such action would have taken
Akroinon. An Arab march so far into Paphlagonia would have allowed Leo
much more time to gather troops than would an Arab attack upon eastern
Cappadokia.llT Thus the great campaigns of the Arabs were in a way much
more vulnerable than the comparatively minor border raids, which would
troops.lis
The smaller scale local raids were also vuinerable to carefuily timed
Arabic frontier defence to the second battle of Akroinon in 739-40, four years
saw the destruction of at least one Arab detach-ment. To th-is can be added the
and.739-40) and the losses incurred during the siege of Nikaia. Further, Hisham
is recorded as: " . . . squandering many of his men . . " ir-t Romania in 724-5'11e
Finally, on one occasion in73L-2, a Muslim force was assembled at the border
and did not even venture into Byzantine territory.l20 This would have cost the
When all the events are examined together it may be seen that the period
from 720 to740 was a difficult one for the raiding armies of Is1am, especially as
compared to the years prior to the siege of Constantinople. If the brief phase of
Byzantine aggression from 71.8 to720 is included, it becomes aPParent that only
11e
Theophanes, g5.
t20
Tabari, Btankinship ,96;Yaqubi has the date as 730-1, Brooks, THS 18 (1898) 200.
67
loss. During Leo's reign three towns were taken from the Caliphate and at least
resurgence. Against these setbacks the Catiphs could point out that their armies
succeeded in taking at least ten named towns and three unnamed fortresses and
had defeated two Byzantine field armies. In addition, while the Nikaia
campaign was undoubtedly a costly one for both sides, the losses of the attackers
may have been offset to some degree by what Theophanes calls tLle: " . . ' large
body of prisoners and booty," they carried away.121 The first battle of Akroinon
also resulted in the falt of the outer city to the besiegers and may have provided
reign the Emperor followed several principles regarding the Arab raids. After
720 he became aware that the Empire could engage the Caliphate in only a
'limited' war if it were to entertain any hope of success. Even at full strength the
thematic forces could not prevent the loss of border-citadels if the Muslims
launched a significant attack. But they could harass the raiders whenever
121
Theophanes, 98. Although Nikephorus'less alarmist comments on the siege
should not be forgotten; Mango, 131.
1" T aban, Blankinship, 97.
62
Most obvious is the military advantage of reducing the opponents' numbers, but
there are others. When the chance of being ambushed existed the number of
raiding parties would almost certainly be lower in order to increase the strength
of each party. Larger units move more slowly and are more easily seen and
would rapidly affect the main arrrryt forcing commanders to bring alarger, more
Otherwise the raid would inevitably be cut short as provisions dwindled and
The knowledge that an enemy force was nearby and preying upon the
the enemy was located or at least its size ascertained, only a rash commander
would allow his looters to range so far afield as to be unable to rush to the aid of
the parent body. The actual quantify of troops assigned to raiding would
probably also be reduced so that a sudden attack would not find the invading
territory and, walled Arab towns in the valieys.læ Most of the Byzantine citadels
were also local military and adminístrative centres, often with an attached
village or town. The town was un-fortified but the inhabitants looked for safety
in the castle itself, which was relatively safe from assault. Thus, with warning,
much of the population and its movable wealth would be ensconced in the
fortress for the duration of an Arab army's stay in the region. The Byzantine
military system on the border was developed to minimize the damage done by
Muslim invaders. In the case of raids too great to fend off, either wholly or in
detail, the unit was able to retreat with much of the population of the region into
a mountain citadel and let the raiders loot what remained. While the
likely that many of the cities listed as captured by the various authorities for the
period suffered damage primarily to their u¡Jortified portions, while the walled
it must be remembered that the experience of almost annual raids was not new
to the area.12s Haldon notes that even before the Arab raids into eastern
12,
For an illuminating description of life on both sides of the frontier, see Haldon
and Kennedy, "The Arab-Byzantine Frontier in the Eighth and Ninth Centuries," pøssim.
The instances of Nikaia (727-8) and Akroin on (732-3) are indicative of this
t24
Herakleios' time (610-41). Thus the situation as it existed on the border in Leo
trI's day had. already obtained for nearly a century. Far from being disruptive
unusual events, by the early eighth century the Arab raids had probably become
a part of the locat equilibrium. By the 720s andlater the lifestyle of eastern Asia
Minor would either have adapted to the arrrual incusions or ceased to exist;
That the raiders were probably not getting rich from the spoils of the
annual forays is borne out by the records of Muslim Pay-scales for the soldiers of
the various regions. Haldon points out that the rates of pay are much gteater
than those recorded for Islamic soldiers elsewhere.126 It is important also to note
that this same high pay-rate suggests that the government of the Caliphate could
not recruit enough men-at-arms for the region using normal methods. In all
likelihood this must be either because the duty was more hazardous or because
Obviously, the reign of Leo III was one filled with military action along its
eastern border. As has been seen, the Empire was able adequately to defend
For the Abbassid period, Haldon and Kennedy suggest: "It can definitely be
126
said that rates of pay [on the frontier] were subsiantially higher than the 80 dirhams
per month whidiwâsusual elsewhere, which suggests thatconditions were arduous
a¡d there was a need to offer men inducements to serve in the area. "The Arab-Byzantine
Frontier," 112.
65
itself d.ue to a policy of ambush and counter-raid when deating with invaders.
But if these actions had been used in isolation, they undoubtedly would have
met with much reduced success. In reality, Leo combined his military solution
around. the Empire to act in concert with him he was able effectively to lessen the
amount of pressure that the Caliphate could bring to bear uPon the Empire. It is
Byzantine fortresses controlled the Black Sea coast, their churches graced local
settlements, and their influence was present in the imperial titles sported by the
local nobility."'
The Arabs certainly understood the need to exert influence in this region.
During Leo's career, virtualty every year expeditions were launched by the
Arabs against different tribes of the Caucasus and Black Sea area with an eye
ttt W.E.D. Allen, Histor)¿ of the Georgian People, Routledge and Keegan Paul, London,
1971,, 80; see also S. Der Nersessian, Armenia and the B)¡zantine EmPire, Harvard
University Press, Cambridge, Mass. 1947,7.
66
Long before Leo became Emperor, he was made aware both of the
advantages of alliance and of the importance of the various peoples beyond the
during the second reign of ]ustinian II (705-11). At this time Leo became
intimately acquainted with both the Alans and the Abasgians. Indeed, if
Theophanes has the details of the story correct Leo left a lasting positive
impression upon the Alans and reaffirmed their alliance with the Empire. [r
addition, he was able to press the ruler of the fortress of Sideron, as well as the
were also convinced to return to the imperial fold after suffering devastation at
the hands of the Alans, who were acting on the behalf of Leo.l2e Thus Leo
shored up the faltering Transcaucasian situatiory but it was not to last. By 771',
128
ibid,80.
t2e
See Map for the geographical setting of the Caucasus region'
67
the Prince of Abasgia as well as Smbat, ruler of A¡menia were once more within
the orbit of the Caliphate. Regardless of the outcome of this instance however,
Leo seems clearly to have tmderstood the necessity of good relations and
generally are not found in the major soufces for the period. This may be due, as
As has been seen, by the beginning of the eighth century, the Caliphate
limited the Empire's ability to take direct action in Fourth Armenia, but
Through the Pontic Sea,Byzantium had access to the Abasgians and the
lesser west Georgian peoples. Allen notes that until the end of the eighth
t" S. Der Nessessian, Armenia and the EmPire, Harvard University Press, 1947,7'
132
Martyropolis, Manzikert and Erzerum respectively; Minorsky, Hudud al-Alam, 395.
68
century, Byzantine control over the Abasgians was reasonably secure'1æ Gero
provides evídence to corroborate this assertion. A letter which Gero says refers
to events no later ttlan736 (the prince Mir addressed in the letter died in
736)
the
congratulates the princes Mir and Arcil for their staunch defence against
attacks of the Caliph's forces. The preface to the letter adequately demonstrates
The text would aPPear to preserve an aPpeal by Mir and Arcil to the Emperor
Leo for salvation from Arab aggresion. Leo rewards them with the symbols of
their clientship and makes vague promises of aid. Nothing is known which
directly relates to any military salvation for the princes, though perhaps the
It has already been seen that, in the early years of the eighth century, the
Alans were favorably disposed toward the Empire, in part due to the efforts of
against them and imposed the jizyahupon them.136 Yet by 735-6, Tabari relates
that the future Caliph Marwan captured tluee fortresses in Alan lands'13t
Apparently the Alans had shifted away from Muslim control at some point over
the intervening ten years. No source gives the exact reason for this shift, but
make the Alans attempt to tluow off their Arab yoke before it became
impossibte. Several campaigns in the recent past saw Islamic forces conquering
the lesser tribes of the region. As early as the reign of Yazidn (7204),Tl:re
13s
Tabari, Powers, L64;Yaqubi, Brooks, THS 18 (1898) 797. Gerc also notes that
another letter to [æo, Ruler of Àphazeti, is preserved by the chronicle' He says that
the text is obviously false, 151. Vet the correspondence itself need not be doubted,
only the substance of it.
1s6
Wahis the tax or tribute imposed by the Muslims upon non-believers under
their control, Tabari, BlankinshiP, 8-. A year earlier , al-Haiiaj was able to pass through
Alania and progress as far as Balanjar in Khazaria' So apparentiy the Alan posiiion
toward the Caliphate was not overly hostile even then'
d.omination. úr 72g-30,a1-]aruah crossed the Kur and the Samur rivers into
Ifttazaria,but on the way he took Hamzþ Ghumik and Shakki, all near
neighhors of the Alans.138 The next year al-]arrah and most of his command
were slain by the l(razars on his return to Ardabil. But shortly thereafter
Khazars and conquered Khrai zan, a!-!Ibal, Sharwan, Lifan, Tabarsaran, Filan,
and settled it
|arshan, and Muskat. He then besieged and took the city of Al-Bab
with 24,000 Syrians.l3e Tabari records Marwan defeating the Tumanshah once
Sindirah, while Ishaqb. Muslim al-Uqayli again seized the lands of the
carried the war into Khazar territory.tnt Lr the next season Marwan led another
t38
Baladhuri,323' Minorsky, Histor)¡ of Sharvan and Darband' identifies Hamzin
as the Humfi, and the Ghumik'as the Tuman of Baladhuri' Hamzin may be identical
to the Hazmadan also mentioned by Baladhuri,322; see Map'
obviously tra¿ gãtadhuri as a so.rtce, has details which must have come from elsewhere'
The BCraLan may be tCrazai, just south of the Samur. Al-Jibal, Liran and Jarshan are
unidentified; see Map.
tno
Tabari, Blankinship, 111.
1a1
Tabari, Blankinshi p, 1.67 . Blankinship identifies Sindirah with the Sideroun
mentioned by Theophanes for the same year (p.103) but states that the site cannot be
identified further. Sidetorrtt is also mentioned by Theophanes during Leo's
77
campaign into the same region and beyond to the country of the Lord of the
Golden Throne:
tt.e region. Yet there were definite limits to its Power. While it was apparently
possible to traverse Alania to raid beyond, Marwan was aware that there was
no
way in which he could retain direct control over the Sarir' A powerful
permanent Arab Presence was simply too fat away to establish the Sarir as any
sort of a client state. Marwan had to settle for a simple demand for tribute;
Similarly, aiízyahwas required of the Alans, but there was no way to retain
Transcaucasian mission prior to his assnmption of the purple (p 87)' tvtrlorl\Y, Histor)¡
of
Sharvan and Darband. dãscribes a people called the Shandan/Sindan,103. Their location
matches with the details which *uy u. gleaned from Theophanes; see Map. TheoPhanes'
reference for73l-Bmistakenly phcôs Sid.eroun within Romania. In the context of the
earlier passage, Sideroun is o-büiously east of the Lazfüans (Arabic' al-Lakz)'
t4 Tabari, Hillenbran d.,34. The lord of the Golden Throne or Sahib Sarir ad-dhahab
(Ar.) is the Khan of the remnants of the Avars who settled just north of the Alans after the
the
dissolution of the Avar Kingdom in the second haif of the seventtr century. lGrown as
Sarir (Ar. 'throne'), they areäescribed in the account of Ibn Rustah, who visited the Sarir
prior io 902 4.D.. Ibn ftustah, Les Atouts, Trans. G. Wiet, from Minorsky, Histor)¡ of
Sharvan and Darband, 1'67-8.
72
Even this powerful citadel may not have been enough to secu¡e the area
presumably to hold tlLis strategically vital site.l€ FIowever, Masudi records that
the Alans could muster 30,000 horsemen and the Khazars are noted by al-Balami
and Ibn al-Athir as having assembled that same number for a single
expedition.la Thus the Syrian garrison could not hope to enforce Arab authority
without aid from Damascus. For Byzantium, however, the result is the same:
Alania slipped further from imperial influence even as the Caliphate established
Leo's best known diplomatic manoeuvre is the imperial alliance with the
Khazars. In the early 730s,Leo acquired a Yrhazar princess to wed to his son
existence for some time previous.las The l{razars had been at war intermittently
with the Caliphate for nearly a century and from 7L7 onward Arab-Khazar strife
was only slightly less common than the near annual Muslim raids upon the
143
Baladhuri,325.
It is not intended that these figures be taken as actual values, but they are certainly
144
indicative of the various authorities; opinions of the relative srrengths of the different
forces in the region. ln this regard the figures demonstrate that the 24,000 Syrians were
probably intenãed to act primãrily as a defensive force for the citadel. Masudi, trans.
'Sp'..'g"',436;foral-BalanriandILnal-AthirseeDunlop,Histor}¡oftheTewishKhazars,
62.
Ln732-3 according to Theophanes, 101; Abu Faraj has731', p'110; Gero notes
14s
that while Michael the Syrian haà79t, the Chronicon aci 1234 puts the marriage i'-,
730, Iconoclasm in the Reign of Leo III' 28-9.
There is no recordõf the ter-s of the treaty. It may have been one of mutual
nonaggression or of financial aid to the Khazars.
73
Empire. By the early 730s,the Khazars would almost certainly have been
advantageous.
which the Arabs defeated a Byzantine force and took 700 prisoneÍs.'n' But for
the same year,Duntop notes that al-Balami and Ibn al-Athir relate a major battie
between the Khazars and the Caliphate, also in Armenia, which resulted in a
pr-rnishing loss for the Muslims.lat No further evidence exists for any joint action
by the Empire and the l{razars, but relations were apparently peaceful until the
146
The enhies in the various sources are extremely confusing chronologicaliy, and
no adequate analysis of the Arab Transcaucasian campaigns has been done. ln addition
to the cirronologiád difficutties the sources often use the same name to refer to what are
apparently diffãrent peoples or different names to refer to a single gfou-p' It is not the
itrt""t of tire pïesent*oit to explore this complex problem except as it bears upon the
central issuebf Byzantine-Arab frontier warfare in this period'
t47
Tabari, Powers, 164;Yaqubi, Brooks, THS 18 (1,898) 797'
1aB
Dunlop, History of the Jewish Khazars. 62. Itis not rnrpossi,ble-that I-eo and
the lcragan õf tfte lCrãzars agreed to a co-ordinated invasion of Muslim Armenia.
The timñg coincides with thé completion of Umar's reinforcement of Cilicia against
Byzantine"attack. Perhaps Leo saoi an opportunity for further exploits in a new theatre of
operations. At any rate,ìt appears that thã venture was a failure for the Byzantines. This
setback may have'been instümentai in the formation of Leo's defensive policy of
ambushing invaders (which was adopted at this time).Conversely, it may be simply that
læo was altempting to seduce one ofìhe local Armenian potentates away from their Arab
masters.
74
under discussion (7174n. The KLrazars i¡flicted heavy losses upon the Caliph's
noted attacking f$azar possessions in strength. While this was without doubt
pr-rnishing to the Khazars, it meant that there were considerably fewer troops
for
Leo had gotten the better of the bargain. In fact, Leo could accurately be accused
of leaving his allies to face the aggressions of the Caliphate unaided. There is
some evidence to suggest that he had treated the Abasgians at Karthli similariy
at some time in the past. To the embassy of Mir and Ardacil, Leo replied:
14e
The Arabs made raids upon the Khazars in7234,728-9, and 730-L. It should
be noted that, had these forceib."n available to the Caliphate for use against the Empire,
it would have been an easy logistical mattef to transporlthem to the imperial border
rather than to the region of al--gab (the Caspian Gates). This could have made the
consequences of the"annual raids much more dire for the Empirgt The Arab-Ktrazar
wars
are as irt"cft a boon to the Empire as the Abbasid revolution and its attendant distractions
wouici be two decades iater.
1s0
Gero, Iconoclasm in the Reign of Leo III, 151'
75
The general impression of the passage is that Leo will aid them with little more
than words and crowns. This attitude may have been hard on his allies, but if
Leo saw the Empire's military situation as desperate, then he essentially was
neighbours.lst Leo did not have sufficient men or money to launch campaigns
into Armenia and beyond unless the reward was great.1s2 By not taking action in
aid of his allies, Leo was saved the expense of such a course and was abie to
military force being directed at his beleaguered Empire. After the siege of
Constantinople, Leo was able to organize a brief but successful offensive against
the Caliphate. But he soon came to understand that the Empire could not
Islam and altered his military policy to timit the damage inflicted in this regard.
This policy was so successful that he was able to in-flict several losses upon the
invaders and timit their depredations primarily to the frontier region. To further
limit the forces the enemy could bring against the Empire, Leo sought out
alliances with others who stood in the way of Arab attack. By creating such
alliances he was able to split the attention of the caliphs and force them to
Since there is no way of knowing læo's actual motivation, it is also possible that he
151
had no intention of ever coming to the aid of any ally unless it suited him.
If he had already lost an anny on one such attemptln720-'1., he would have been
1s2
commit men and resources to actions other than those which targeted the
Byzantines. There were no territorial gains for the Empire during Leo's reign.
Nor did the Caliphate suffer any major immediate consequence of their setbacks
at Leo's hands. The worth of Leo's poticy was not in that which it accomplished
as much as it was in that which it prevented and that for which it provided.
Through careful application of martial power and negotiation, Leo preserved the
Empire and reestablished the Arab-Byzantine frontier. It took many years for
the realization of this policy, but it eventually allowed for a further resurgence of
Byzantine military fortune under Leo's son and successoÍ, Constantine V.ls
Constantine was able to take the rejuvenated Empire of his father and,
capitalizing on events around him even as Leo had before him, take the border
It is worth bearing in mind that the last actual dynastic succession was over half a
,5,
century before when Juãtinian II succeeded his father Constantine IV in 685-6' From the
deposition oÍ justinian ifi 695-6, the irrrperial office became the province-of whomever
.oùt¿ find sufficient backing in the militaryto take it. læo himself seized power in just this
manner. With the accession-of Constantine V came the reestablishment of imperial
succession and the stability which this sometimes entails.
77
VI
CONSTANTINE V
days into his reign, Constantine took his army into the Opsikion
theme to
neither Byzantine nor Arab soufces provide any greater detail'is For the
same
The
year Baladhuri records an unsuccessful Byzantine attack upon Malatyah'
to
the Arab border. It was no longer necessary for the Byzantines simply
tuu
Baladhuri,29l.
78
the Empire to take the offensive as it had not been under his father - making
raids into Arab lands and looting Arab towns, just as the forces of the
But this new policy could not be more fully enacted until Constantine
Opsikion and lifetong supporter of Leo III, rebelled against Constantine and
had himself proclaimed emperor. The civil war which followed lasted until
the late fall of 743 andoccupied much of the attention of both contestants for
its duration.lsT
The Arabs took advantage of the internicine stuggle to raid the Empire.
According to Theophanes: "When the Arabs leamed of the civil war between
these men they took many prisoners in Romania, Suleiman was their
general."is8 Tabari confirms Theophanes testimony but adds that Leo III, at the
156
1s8 Theophanes,106.
1se Tabari, Hillenbrand, 68. Yaqubi also preserves this tale but adds that Artavasdas
was beside Leo and that there wãs no engagement before the Arabs withdrew;
79
Artavasdas. Yet it may be worth noting that only Theophanes credits the raids
of this period with significant consequences. The Muslim soulces are united
in
record-ing simply the occurence of a raid and its leader. The impression to be
gathered from the Arab authorities is that the raids were not major' Overall,
it
would appear that, even when divided against itself, the Byzantine military
was able to offer effective resistance against aggression.lto While this statement
major Byzantine offensive upon the cessation of the civil war. Even after the
losses which must have occurred at that time, the imperial army was capable
of immediate action.161
Caliph Hisham (72443). At some point during his reign he fortified a large
160 It should be noted that the Muslim sources do not record severe internal
difficulties for the Caliphate until 743 and the death of Hisham. Thus there were no
divisions within the Mislim world to distract the sources or the raiding forces which
they record.
161 There is no mention of Byzantine opposition to the Arabs Íot 742-3. However, if
the thesis previously advancád concet"úñg the comparative unassailability of many of
the Byzaniine citadéts is correct, the Arabiaiders w-ould have needed considerable
forceio affect any Byzantine population centre. Thus even when unopposed, the
invaders would often be limited to Iight raiding rather than conquest'
80
were carried out. If the period of fortification did not span the length of
Flisham's reign, and was enacted over a relatively short period, the most
logical time would have been in his last few years. As has been seen above,
while Leo's armies were definitely troublesome to Muslim raiders, they posed
years of the reign of Leo, his successor, Constantine, was of an age to take an
active role in the exercise of imperial power. Indeed, the young emPeror must
have been active in the military prior to his accession to have acheived the
early successes that even Theophanes credits him with- If, as is probable,
Constantine d-isplayed the same aggresive style as junior emPeror that he did
upon taking the throne, there was ample reason for an astute caliph to begin a
fortification program on the Byzantine border. Thus, the defences were most
162
Baladh¡¡1ri,2SB; Baladhuri also records the construction of another fort, al-
Muttakkab, in Hisham's time, p.258.
Bagas/Pagrae, also knbwn as Birjirik is inland of Antioch. Katargash may be
Karsaga. Murah, Buka and a]-Muttakkab are thus far unidentified.
81
likely to have been und.ertaken in the late 730s or after, possibly as a reaction
to the major loss at Akroinon ín739.1æ It is even possible that Hisham used
the Byzantine civil war to take defensive action as well as to launch offensive
raids.
more fulty upon the easternborder situation. He lost no time in seizing the
and the Arab civil war which followed the death of walid II (reigned 15 April
r6t Other than the abortive seige of Malatyah 1n740-'l.,no overt action took place in
the part of Constantine's reign\Z+t-ZZS¡ tñat overlapped Hishart's (724-43) which
stands out as an obvious imþetus toward such defensive measures. The loss of so
large a force as the one destioyed at Akroinon could have had far-reaching
imþtications for the Arab miliiary. As was noted above, it took several years for the
Caiiph's forces to recover from the spectacular losses incurred in the seige of
Constantinople (716-8).
t* Adata and Zopetrah respectively; Baladhuri,2gg.Ibn al-Athir records the fall of
Zlbatrahunder tire previous year, Lut Baladhuri specifically states that it was attacked
in the days of Waiid; Brooks, (790i0) 202'
lriar*an b. Muhammad is the future Caliph Marwan II (7114-50) who
championed the offspring of the murdered Walid II as a means to acquire the
Calipirate for himseú. Tt ls is the same Marwan who was so active against the
Traiscaucasian peoples throughout the middle decades of the century'
165
Baladhuri, 299; anð.ibn ai-Athir, Brooks, (1900) 202. Both note 'that Zibatrah was
rather weakly rebuilt by the Arabs and once more reduced by the Empire during
Marwan's Cánpnate. Bãhdhuri goes on to remark that this pattem contirues for
Zibaftahinto the days of al-Mamun (813-33)'
82
revolt of the city of Hims against the new Caliph, Marwan T' 7'6 According to
Baladhuri, imperiai armies fell upon the city of Marash and took it on terms
after a seige. Again the cify was destroyed rather than being garrisoned'1ó7
Theophanes records the same raid but adds that the Byzantines continued into
Kibyrraiot navy was at anchor in Cyprus at the time and, taking advantage of
the opportunity, was able to surround the Muslim armada and cut it off ftom
hands in the early part of Marwan's reign, it is not surprising tl-rat he was
1* Hims/Emesa was in revolt for several months beginning in June of 745. See
Baladhuri, 294-5, and Tabari, Williams,4-9'
167
Baladhuri,2gÇ5. Marash/Germanfüeia was the birthplace of Leo III and the
strategos Artavasdas. Theophanes also records that Constantine transferred some of
hir .el-atirr"s from Marash to the capital at this time. The city was rebuilt by Marwan
n747-8; Tabari, Williams, 121; bufit was again sacked by the Empire and once more
rebuilt in the days of al-Mansur (7il-mÐ.
158 Theophanes, 1L2. Doulukia is Duluk in Syria; whether it was g_arris_oned rather
than destioyed is not recorded but the latter is much more likely. See also
Nikephorus, Mango, c.67,p.739,who puts the event n747-8'
16e
Theophanes, 113. Cyprus was at this time a condominium between the Empire
and the Caliphate. I^74i-9, Hisham had sent a force there to compelthe inhabitants
to choose either Arab or Byzantine suzerainty; Tabari, Hillenbrand' 720'
83
well as increasing the defences of Hisn Mansur. IrL addition to upgrading the
fortifications there, Batadhuri notes that Hisn Mansur became the station for:
" . . . a large host of the troops from Syria and Mesopotamia in order to
the pressure of the Abbasids. Marwan fled but was murdered in EgyPt'
was no more and the Abbassids established themselves in the Caliphate under
The new Caliph was as aware of the troublsome Byzantines as had been
lands among them.171 His concern was not misplaced, for Constantine
Constantine then invested Malatyah and, followiog a seige, the citizens were
razed.to the ground by the Greeks, who left nothing but a granary of which
Baladhuri,299.
17r
ibid,257.
84
garrison in the citadel of Karnak.l73 The response of the Caliphate was a raid
From the end of the Malatyah campaign until the accession of Mansur
(g / t0 lune 754) there is no record of any border strife between Empire and
Caliphate. Yaqubi states that Salih b. Ali's raid into Byzantine territory n754-
5 was the first since 7424.17s Tabari holds that the leader of this year's
expedition was Abdalahb. Ali, but that the army did not succeed in
772 ibid,2g1. He also suggests that Constantine knew that no aid would be
forttrcomiog to defend Mäiatyah. Constantine is said to have taken advantage of the
intemal strife within the Caliphate following the rise of the Abbasids to make his
attack.
Nikephorus, Mango, c.70,p.7[3,confirms the attack on Melitene; as do Yaqubi,
Brooks,(1900¡ZtZ,a¡d AÈu Faraj, who lists the Byzantine commander as: "Ashkirash,
captain of the Armaniko . . l',p.11'2.
r73 For Hisn Kalidiyah/Claudias (a fortress near Melitene), see Baladhuri, 291', artd
Abu Faraj, 113. For kahkala/Theodosiolis/Erzerum see Baladhuri,372, Theophanes,
116, and Âbu Fara;, 1L3, who also refers to Kamak, but places the events in 755 A.D.
The attack on Kalikala was reported by Baladhuri as having been led by one
Kusan al-Armani, who may be identical to Abu Faraj's Ashkirash of the Armaniko.
As such he would probably have been the commander of a sizeable detachment from
Constantine's main force it Malatyah. Kalikala was also:" . . . razed to the ground;"
Baladhuri,312.
17a Tabari, Williams, xxvü, 197. No details are given for the raid, and no other source
mentions it at all.
17s Yaqubi, (1900) 792. Ofcourse this contradicts Tabari's account for750-1, but the
generaltimpression for the decade 743-753 is one of inactivity and weakness on the
part of the Caliphate.
The accomplishments of Salih's raid are not known'
85
at the head of 80,000 men r¡nder the following year. The invaders are said to
have turned back upon hearing that Constantine was taking the fetl'd'ln
Perhaps there were two raids Íor 754-5, one before and one after the accession
of al-Mansur. But it seems mofe likely that Yaqubi is mistaken here, for
Theophanes and Tabari seem to match each other closely in most of the
details.
Mansur's reign may not have troubled the Caliph as he seems to have been
more concerned about the state of the border citadels than upon any offensive
Various sources record a major reconstruction effort put into effect by al-
would grow to a major Arab border cify over the next century. Finally,
Baladhuri notes that Marash was rebuilt at some time during the reign of al-
Mansut.iso
The details available for certain of the fortification efforts suggest that
they were quite extensive. Marash was strengthened even though it had been
Mansur raised the walls and. increased the garrison by 400 men.182
1n757-8,a1-Mansur sent Salihb. Ali to begin work on the city which had once
been a major population center and military staging point. 70,000 workmen
and troops were gathered for the task, and when it was completed the walls
t80 Adanah - Baladhuri,260;for Adanah's growth and its role in later border
situation, see Haldon and Kennedy,707-8. Marash - Baladhuri,295.
rer ¡4 Z9|;Tabari, Williams, xxvii, records the Marwanid reconstruction, 121'
182 1o addition, Abu-t-Abbas had already raised the number in the garrison by 400 in
750; Baladhuri,257.
87
were defended by 4000 men besides, as Baladhuri puts it: " . . . the necessary
with over 4000 defenders was no longer considered safe from Constantine's
forces.185
upon acquiring in-formation concerning the size of the host arayed against
for the previous year as to merit attention. Salih b. Ali is said to have entered
imperial territory only to retreat upon hearing that Constantine had taken the
field in response. The passages are virtually identical save for the substitution
of the imperial forces for those of the Caliphate in the latter instance. In each,
the enemy is reported to have fled before greater numbers of the defender.
Perhaps what each author has preserved is not similar events for two years/
183 ibid, 293. For the rest of the fortifications there are no details as to their extent.
Kalidiyah is generally treated by the sources as an outpost of Malatyah rather than as
an ind'ependant fortrêss. Concóming Kalikala, Baladhuri notes that the population
-ar.uolomed from the Byzantines and allowed to resettle in the former locale,313.
184 1bid,2g3.
18s However, the Abbasids were able to accomplish their refortification with
comparatively little intervention from the Empire. Since 755-6, Constantine had been
embioiled in a bloody war on the westem frontier against the Bulgars. According to
Theophanes, this conflict was brought on by a series of imperial border fortifications
erected by Constantine in order better to control the Bulgars (Theophanes, 119). The
war in thã west continued irregularly for the rest of Constantine's reign- It is not the
intent of this sbudy to address Ihis cónffct except as it bears uPon the Byzantine/Arab
border situation; for the Bulgar wars, see Theophanes, LL9-135.
185 Baladhuri,293.
88
but prejudiced, views of the occurences of the same season. If this is so, then
both Empire and Caliphate mustered sizable forces for the campaigning period
but each declined to commit them against an equally impressive army. Thus,
Theophanes relates that after a season of stalemate Salih retwned beyond the
passes with: " . . .o.Iy those few Armenians who had gone over to hi*.tt187
With the fortifications completed the Caliphate was once more capable
of pursuing a more aggressive border policy. This policy was certainly made
Large Byzantine forces, probably comprising the majorify of the Tagmata and
at least portions of the westem themes, were often in the field at the opposite
end of imperial teritory. By necessify this would have made any sort of
While Constantine was personally active in the west, the defence of the
east necessarily fell to the various thematic strategoi who came under attack.
Such was the case in 759-60, while the Emperor was campaigning rather
187 Theophanes, 119. This interpretation also coincides with Tabari's record of Salih's
retreat.
There is also the possibility that two distinct years' of raiding are recorded by
the two sources. Such a confusion would have been easy for Theophanes to make.
The year as he records it begins September 1 and ends August 31. Thus Theophanes'
year 757-8 would last from 1 SeptemberTST to 31 August 758. Tabari's coresponding
ônny, following the Arabic calendar with its shorter year, runs from 25 May 757 to 13
May 758.
188 There are two main phases to the Byzanitne/Bulgar conflict. From 758-65,
Constantine was in the field every year save lor 767-2 and763-4. The second phase
began aJter Constantine repudiated the peace (of which there are no details) rr:.772-3
and continued until his death (L4 September,775).
89
the
unsuccessfuily in the Baikans, Theophanes relates a significant defeat of
Yaqubi records two more raids fot 760-1' and762-3 but offers no more
information on either.leO Abu Faraj is more infomative for the latter event:
The words of Abu Faraj raise a difficulty concerning Marash. The last known
18e Theophanes, 119. There are fwo rivers by the name Melas, one in coastal Isauria,
the otherã hibutary of the lower Sangarius. Either is possible, since Paul would
be
far from the A¡ea of the Armeniac thãme in either case. Yaqubi confirms the raidbut
provides no details beyond the name of the commander - al-Abbas; Yaqubi, Brooks
(1900) 733.
t" Baladhuri,295.
90
retaliation by the caliphate, yet there is no other record of any imperial action.
The passage seems to suggest that the city ítself was the target of the Arab
attack, which would lead one to understand that the Byzantines were in
possession of it. Perhaps the reference of Abu Faraj to the 'spies' taken by the
and therefore the city as well. Rather than re-garrison the city, the Arabs
chose to remove their subjects from the district of Marash. A similar case
Yaqubi records that the Arabs carried out trnspectacular raids for the
next several years while Constantine was involved in the Bulgar wars. Tabari
is at variance with Yaqubi's details on more than one occasion. Fot 765-6,
Tabari has Salih b. Ali as the leader of the expedition and notes that it
Ali.les Given the custom of both the Empire and the Caliphate exploiting the
other's weakness in order to raid over the Taurus border, it is logical to expect
1s3 It appears that this is an instance of Byzantine conquest that went unrecorded by
aoy untiho.ity save perhaps for the vague reference of Abu Faraj above. It will be
seen that there are other such 'hidden'raids'
1ea Dabiq is: "An important Muslim base on the Byzantine frontier, north of Aleppo;"
Tabari, Kênnedy, 40. Dabiq is possibly to be identified with Duluk'
1es Little can be said'¡¡ith any degree of certainty conceming such contradictions.
Perhaps Salih sent his son on a minor expedition. Perhaps a Byzantine aflny was in
the fietd and Salih held his force in res.Ñe in case of need. ln aoy event the season's
raiding appears to have been of little consequence.
9t
major assaults to have been launched by the Arabs while Constantine and the
Tagmata were occupied in the west. Indeed tlLis would seem to have been the
case when the strategos Paul was kitled 1n759-60. Yet for the next several
years,as has been seen above, only minor raids are recorded. This is due to
Muslim-held Armenia at this time. Tabari states that extra troops wefe
stationed at Tiflis due to concern about a potential revolt, but unless the
Bab al-Lan had come under Muslim domination at some time du¡ing the
Bab al-Lan was garrisoned with a guard of Arab cavalry.lÐ These factors do
let Abu Faraj has 762 A.D.,p.774. Theophanes places the assault 1n763-4, but says
that a second Khazar expedìtion was made in the following year,l?3. Tabari
preserves the instance oi the Turkish attack n764-5; Kennedy, 1'4-5. -- -
^ Atthough the date of the attack is certainly open to dispute, all the atcounts
agïee on the r.ãpe of the invasion and the damage inflicted upon the populace in the
area. Tabari even states that Tiflis fell.
1e7 Tabari, Kennedy,40. Theophanes definitely states that there was a bloody battle
between the Turks and the Arabs,1.24.
nothing to indicate that there was any apparent weakness on the part of the
Caliphate which the Khazars might exploit. The situation does not seem to
Since Constantine initiated the war with the Bulgars and even prolonged it
confident of his abilify not only to defeat the Bulgars, but also to repel the
that the Muslim raids would necessarily be minor. He may have arranged this
the defeat of the Armeniacs under Paul. Certainly Constantine had seen
Regardless of the instigation of the Khazar attack, its effect was highly
beneficial to the Empire. Apart from the disruption of the region which must
have occurred, the Khazars were attacking the very territory which was home
to the raiders sent by the Catiphate to loot Byzantine lands. It would have
2oo Theophanes,62.
201 See above for similar actions taken by Leo IIi, see n.77, andp'69,74-6' Given the
bias of the surviving sources against Constantine, it is hardly surprising that no
information on his diplomatic efforts is extant.
93
been a rash commander indeed who sent out a large raiding expedition only to
leave his base exposed to the marauding Khazars. The raiding forces are the
invasion. This would obviousty limit the numbers available to plunder across
the Taurus. Thus, there is little wonder that there were few raids of note in
this period.
With the Khazar threat hotding the Arabs on the eastern border in
against the Bulgars in the sulnmer of.768.202 While the emperor and his
According to Tabari after a fierce seige the town was taken by storm. Yet for
the same event, Theophanes states: " . . . Abd Allah beseiged Kamachon all
anytlLing."2M The following year another raid was made under Yazidb. Usaid,
but Tabari notes that: . . . he did not lead the people to the land of the enemy
202 There is no record of a peace tteaty,but hostilities ceased. The summet of 766
saw Constantine lose a largè fleet of troop ships in the Black Sea; Theophanes, 126'
203 Kamacha-Ani or Kamachon, Baladhuri,288; Tabari, Kennedy,42; Theophanes,
132. The last reference to Kamkh was in 751 when Constantine th¡eatened it during
the Malatyah campaign, Baladhuri,397. Yet there is no indication that the city fell at
that time. Either eotlitantine succeeded in taking Kamkh n751' or it fell to another
unrecorded Byzantine assault as in the case of Marash.
lncidently, Baladhuri notes that Kamkh remained hotly contested and changed
hands repeatedly over the next cen|ury and more; Baladhuri, pøssim.
2oa Theophanes, 132. Little can be deduced from such completely contradictory
passages, save that perhaps they refer to events from different years.
94
but stayed at Marj Dabiq."20s Two more years of unimpressive raids foliowed,
and for 769-70, Tabari again relates that the raiders did not go through the
although it saw only a minor raid, found the Emperor suing for peace with
Mansur.2o7 Yet, as damaging as the attack on Laodikeia may have been, there
is nothing to suggest that the attack alone should have convinced Constantine
20s Tabari, Kennedy, 50. Yaqubi also records the raid but for the previous year/
Brooks, (1900) 733.
207 Lbid.70.
95
other source and may well exaggerate or misrepresent the situation in any
report such success. Yet if the essentials of the record are accepted - i.e. that
an Arab expedition beseiged Sykes, eluded the Byzantine forces sent against it,
and then inflicted a defeat upon the imperial forces massed there - then it
becomes easier to understand the motivation for Constantine's bid for Peace.
conflict at all. Since the cessation of conflict against the Bulgars, The Byzantine
mititary appeafs to have relied uPon the thematic troops almost entirely.
There is no record of any action taken by the tagmata on the Taurus border or
elsewhere.2æ The Arab raids on Kamkh and Laodikeia were resisted only by
208 Theophanes,133.
2oe There was, however, a Byzantine offensive launched against Muslim Armenia in
770-1,. it was in atl likelihood a minor raid carried out by local thematic troops; ibid,
132.
96
the garrisons of those cities. As has been seen above the troops called upon to
relieve Sykes were all thematic. Concerning Kamkh and Laodikeia, it is true
that the Emperor and the tagmata were already committed against the Bulgars,
but Constantine was obviously aware of the attack on Sykes early enough to
take part in its defense. He did so without the use of the tagmata.
In his entry for 755-6, Theophanes records the loss of a fleet of 2600
troop ships in a fierce Black Sea storm. The number is no doubt exaggetated,
but the damage to imperial forces appears to have been significant. In all
probability, the tagmata (and possibly some thematic troops) would have taken
ship to Bulgar lands. If such is the case, Constantine may well have lost the
majority of his most seasoned troops to the Euxine tides. This potential
crippling of the tagmata would limit Constantine to just the sort of action that
Constantine would have been left with littte choice but to explore non-martial
options to safeguard his border.210 Mansur's reply to the request is not known,
2r0 Of course, paying off invaders rather than opposing them militarily is a,device
which dates baãk before the fall of the westem Roman Empire. Economically, paying
tribute could well have been more cost-effective than fielding a defense; certainly this
is the case if the defense is known to be inadequate. Finally, given Constantine's
behavior in his negotiations with the Bulgars, it is overwhelmingly likely that he
would have adherèd to the peace only for as ong as it suited his pulpose. For his
breaking of the Bulgar treaties, see Theophanes, 119, 725,1'34.
97
upon by the inhabitants of the city and one thousand Arabs are said to have
been s1ain.211 Yazid.b. Usaid surrounded an unnamed fortress and looted the
resistance and broke off the raid, presumably to refu¡n home.213 The
tagmata could never have reached Hadath in time to stop the invaders. They
could only have arrived in time to meet the A¡abs deeper in imperial territory.
Also, the majority of the surviving tagmata would probably have been serving
Constantine V died. Less than one month later, Mansur, who had been Caliph
for more than half of Constantine's reign, also died. His son and successor, al-
the Taurus border, thus continuing the tradition of increased Arab border
defenses begun over three decades before by Hisham. That al-Mahdi saw the
made the border into a no-man's-land rather than simply an Arab region
the Arab situation reveals a consistent policy for the Taurus border. To judge
offensive system of defence on all fronts but nowhere was this more evident
than in the east. Attacks launched by the Byzantines on Arab territory could
realize several benefits beyond the obvious one of a victory on the field.
redress the balance of power on the frontier. Economically, any booty could
several border towns and forts must be assessed for its effect upon the balance
of power. During his reign, the Emperor captured ten Arab frontier towns.
Of these, six are recorded as having been destroyed rather than garrisoned,
one is said to have been occupied and one reported as having been razed and
in the region.21s The above actions go far toward demonstrating the existence
Several sound justifications exist for this policy and therefore the abondonment
defensive site has been brought into question. Also, the expenses involved in
the refortification and the establishment of a garrison are also likely to have
been sizable. Yet, the most prohibitive factor must have been the logistical
possibility in the case of a large body, as would have been necessary to hold a
imperial shipment of supplies. Local produce on the frontier would have been
r¡rueliable and as subject to Arab attack as the site had been previously.
be seen from the grain fleets brought from Africa to the Imperial City in the
years prior to the loss of the Carthage to the Muslims. But the expense was in
atl probability far too high to merit a similar scheme for a series of bord.er
apparent that they did not suit the Emperor's PurPose. Flowever, to
Haldon and Kennedy have descibed the style of settlement in use by the
216 Hendy has already demonstrated the extremely local nafure of the Byzantine
economy in the east during this period. M. Hendy, Studies in the Byzantine
Monetarv Economv.300 557-9.
+
101
Generally then, the frontier consisted of Arab walled trading centres in the
defensibillty."t
becomes clear. It was undesirable, and in fact impossible, for him to hold the
captured towns without committing far more men than he could spare. The
217 Haldon and Kennedy,l}l; Haldon and Kennedy also note the anonymous Hudud
al-Alam which describes the standard Byzantine settlement pattern: "Most of the
districts are prosperous and pleasant, and have (each) an extremely strong fortress, on
account of the frequency of the raids which the fighters of the faith direct upon them.
To each village apertains a castle, where in time offlight (they may take shelter)";
Minorskv. Hudud. 156-7.
2ts ibid.1o9.
invasion routes, it was possible for the border troops to become aware of an
impending attack in time to notify the local thematic headquarters. For such a
economically and militarily, to allow the captured sites to fall into ruin rather
than attempt to man them. In this wãf r the Caliphate suffered not only the
expense of the initial loss, but also the additional costs involved in refortifying
policy of aggresive defense and assessed the merits of such a scheme; it now
one way of doing so is to analyse the behavior of the Caliphate. As the focus
of Constantine's works, the doings of the Muslims can indicate the degree to
220 Itshould be noted that this is precisely the system described in the 6th- and 10th-
century manuals on border warfare translated by Dennis; 6th-century ,23-9;1'0th-
century,153.
It is also significant that such aforce, in addition to being cheaper to maintain,
would also have been a less easy and certainly less profitable prize than a rather
loosely held town that also served as the hub of the local economy.
221 One need only refer to the preceeding pages to find that, for this period, there
seemed to have been little question of allowing the ruined sites to ¡emain so.
103
forces of Islam took the Byzantine atiacks very seriously indeed. Every Caliph
money and effort going into defensive works, obviously there would be less
support available for offensive actions. Such support was usually necessary/
for, as Haldon and Kennedy point out, the annual raids on imperial territory
The effects of the Byzantine attacks may also be seen in the relative
weakness of the Arab raids of this period. For the entirety of Constantine's
the ten towns and citadels taken by the Emperor. In additiory it is necessary
to consider the nature of the fortifications of the Empire and the Caliphate
essentially unprotected suburb outside its walls. The general mode of defense
when attack was imminent was to move all inhabitants of the suburban area/
along with any portable wealth, within the safety of the castle walls. While
the overburdened facility may not have been able to sustain the additional
supply and provision the troops neccesary to besiege such a citadel and hold
It is obvious that any attack against such a settlement would involve the
looting of the unprotected and abandoned buildings outside the walls. This
would, in alt likelihood, have gone on regard,less of the success, failure or even
existence of an assault on the citadel itself. But the booty acquired from such
'conquests' must have been very small and the disruption of the local lifestyle
may well have been equally minor. In such an unstable environment, with the
only succor available inside the keep, most things of value would be lost
danger. Further, if the locals were unable to transport some object of value
due to its immovability, then the Arabs may well have passed it over as it
possibilify that at least some of the Arab 'conquests' of the period were, in fact,
only partially successful in the manner described above. This appears to have
been the case in the instance of Kamkh 1n766-7.n4 htis quite possible that this
223 If the relief force did not catch the raiders before the passes, they could easily
have gotten into position to ambush them during this most vulnerable stage of the
joumey. It would have been dangerous to slow the retum in such circumstances. See
Dennis, 233, for the imperial fondness for this type of attack.
22n Baladhuri,288;
Tabari, Kennedy, 42; Theophanes, 132. See p. 107, n.203. Sirnilar
partially successful raids appear to have occurred at Ikonion and Akroinon in the
reign of Leo; see n.89 above, for lkonion. For Akroinos, see n.108.
L05
seige of Sykes.zs
If the style of Byzantine settlement on the frontier means that the targets
of Arab raids may not have been seriously disturbed by them, th" ,u*. *uy
not be said in reverse. Given the Arab settlement-type of walled-town, the
same defensive options were not available to those within. A walled city
might provide a more difficult target initially, but this is debatable because the
the walls were takery the city as a whole, which was generally of some size,
was open to the invaders. At this point, even if there were a central strong-
hold within the city, it was probably unequal to the task of housing the fleeing
citizens and their possessions. Nor did the strong-hold have any more
advantage of natural defensibility than did the city itself. Thus, whether the
central strong-hold fetl or not, Byzantine conquests are likely to have been
commercial centres and home to merchants, the Arab city was certainly a
226 Haldon and Ken¡edy state: "A few general points can be made in connection with
these [frontier Muslim] settlements. The fust is that they were all in the plains, on
fertile sites beside rivers, and the Arab geographers are lyrical about the richness and
fertility of many of them, especially in Cilicia;" Haldon and Kennedy, "The Arab-
Byzantine Frontier," 109.
1.06
richer púze than a small Byzantine border installation, which even the Muslim
on the Taurus frontier was extraordinary. The number of conquests tells only
part of the story, for his captures were also richer than those of his adversaries.
Between the original losses and the expense of refortification it is probable that
only possible remnant of the emperor's diplomatic efforts, the invasion of the
coincidence.
pursuit of the Bulgar conflict. He had only limited success in the west and the
concentration of time, money and troops elsewhere allowed the Caliphate the
unmolested; it is no mere coincidence that this period saw the most damaging
227 Rather they use theword 'hisn' (fortress) or'qila' (castle). Haldon and Kennedy,
"The Arab-Byzantine Fr ontier," 9 6-7 .
zze should be remembered that, in addition to the acfual cost of the construction,
1¡
the Muslims also alloted to their fighting men stipends for their service. The stipends
for this frontier were greater even than normal; Haldon and Kennedy, "The Arab-
Byzantine Frontier," 112.
L07
Yet it is important to recall that even these raids, certainly the worst
wrought by Arab forces in the previous century, or even with the damage
inflicted. during the time of Leo ltr. The defenses of the Empire, even though
the Black Sea, appear to have been capable of bearing up under the might that
As has been seen, over the course of Constantine's rule, the fortunes of
Emperor was able to build upon the foundation established by his father, Leo
III. Constantine also appears to have possessed Leo's ability to discern the
possible from that which was beyond his grasp. ]ust as Leo saw that his
forces could not bear up under the strain of a reign of constant border warfare,
Constantine saw that his own Empire was not capable of a war of reconquest.
22e Obviously the state of the Caliphate also must be taken into account. The
Abbasids, evén after the revolution which put them in power, faced significant revolts
and challenges to their regime; this must have diverted considerable attention from
the western frontier. Yet, it is apparent above that there were far more factors
contributing to the redress of the balance of power than the intemal difficulties of the
Caliphate can claim.
108
VII
CONCLUSION
assess the value of the imperial response to Arab attack. It can now be seen
that no such analysis would be possible without the inclusion of the non-
and, although they occasionally raise more questions than they answer, it is
these texts which contain the necessary data to check the testimony of
It is also obvious that there is more to be done on thê subject than has
been attempted here. There are other Arabic sources, more peripheral to this
study than Tabari, but of value nonetheless. Until these texts are made
Ffowever, while the current work may not be complete, it can contribute
certain insights into the military situation on the Taurus border in the eighth
century.
It quickly becomes apparent that Bury was not far off in his general
glean more detail from the Arab sources, detail which allows for a more
specific understanding of the tactics and practices of the imperial forces. This
1.09
detail shows that both Leo and Constantine had capable military minds and
were able to separate the practical from the tmattainable. Each seems to have
father and son perceived correctly that the Empire need fear for its existence
against the might of their neighbor to the east. Should the Caliphate ever
tfuow all of its power to the west of the Tau¡us, all other imperial concerns
In fact, the invasion of 715-78 was something not far from such a
situation. Leo's ability in blunting the thrust and in severely punishing the
forces of Islam at this time have long been recognised and need not be dwelt
upon. But the impact the invasion may have had uPon the Empire and the
new Emperor should not be overlooked. From that time f.orwatd, Leo made
every effort to be certain that the Caliphate never again gained the opportunity
It would have been difficult for Leo to have escaped the conclusion that
the Empire only narrowly avoided annihilation at the hands of the Muslims in
alternately damaging the already weakened Arab border zone and fortifying
his own defenses. As the Caliphate recovered, attack no longer was a viable
option for Leo. hstead, he concentrated upon the Empire's relatively meager
impossible for the Emperor's armies to prevail if Leo had launched an attack
110
against the entirety of the invaders. Yet, in order to raid and loot effectively,
the Arab forces commonly separated shortly after entry into imperial territory.
By following aByzarttine tactical dictum in use two centuries before and after
his time, Leo was able to inflict small scale but widespread damage on the
raiders. In due course, Leo pushed back the zone of conflict to the region
bordering the Taurus. Only rarely did the Arabs penetrate beyond the
they did, the Emperor was able to respond with the necessafy troops to
confront the enemy. The long trek though Asia Minor gave Leo adequate time
to muster the themes near the capital to supplement the Tagmata. The results
obliterate the raiding tendrils of the main Arab host, Leo was able to safeguard
the bulk of Imperial land from devastation. By allowing the situation in the
border-zone to remain in the same state of chaos in which it had been for a
other enemies of the Catiphate and making coûunon cause, Leo was able to
stirred up resentment of Muslim power among the Khazars, A1ans and others.
11.1.
commit troops and funds to pacification and governance of the area and to the
retention of Arab power in the region. On occasion, it also provided Leo with
timely and much needed aid when sizable Arab armies appeared over the
menace the rear of any Arab invasiory the Emperor held a weaPon against the
Caliphate which could not be countered by pouring greater forces into Asia
Minor.23o
the offensive and extended the zone of conflict to include the Arab border-
protect its own territories from imperial attack. Coupled with the internal
distress in years just prior to and including the Abbasid Revolution, this had
the effect of placing the armies of Islam almost entirely on the defensive.
poorly defended Muslim towns and fortresses under the surveillance of his
This red¡ess in the balance of power on the Taurus was firmly enough
forces to the wars against the Bulgarians on his western border (758-65 A.D
and,772-5). Upon Constantine's death The Empire's eastern border was more
seems to have been wholly mititary. If he gave any thought to the furtherance
the Byzantine armed forces appear to have been equal to the task he set before
the course of his reign because: " . . . he did no lasting damage to either enemy
Tread.gold later states that this was due more to the limitations of the Empire's
ability rather than his own, it seems clear that he has misunderstood
site demonstrates that Constantine did not wish to expand his Empire. He did
want to defend firmty and easily the lands he held. To this end, he set about
destroying the staging points for Arab attacks and thereby creating an
environment hostile to their raiders where before it had been territory readily
The course of eighth century military relations between the Empire and
Byzantines over the years 7L5-775. The fortunes of the Empire went from
adequately defended under Leo ltr. It fell to his son, Constantine V, to take
the rejuvenated Empire of Leo and carried the battle to the Caliphate, securing
Bt n cK t.n
7 r-- --:çå
A?ÍtocHEtAo/-- rxà";g-!5t
El+16óErvÁ¡ff (ÐgORDER.LtÉES'
- o¡f{trq!¡9HAfE
\h&'Þ j,ffi
-.- hÀN AYú¡ffNI ffIqTARY R66S
th\ llERAPoLrg
- --Le$ER ROADS I Ame TiVAS|ON R@tgs
o204060
Mr4l
KMffi
o 2D 40 60 8o too
1.74
Haldory J.F. and Kennedy, FI. "The Arab-Byzantine Frontier in the Eighth and
Ninth Centwies: Military Oryantzalon and Society in the Borderlands,"
Zbornik Radova Vizantinoloskog Instituta.l 9 (1 980) 7 9 -1'1'6.
Sandra Schumann
WriText Communications + Graphics.
-
11.5
VItr
BIBLIOGRAPHY
Primarv
+ Sources:
A-t-Baladhuri, Abu al-Hasan Ahmad ibn Yahya, The Origins of the Islamic
SLaLg trans. P.K. Hitti, Columbia University Press, New York,19L6-24,2
volumes.
Brooks, E.W. "The Arabs in Asia Minor (641"-750) from Arabic Sources,"
Tournal of Hellenic Studies.lS (1898) 182-208.
ibn al-Tiqtaqa, Muhammaci ibn Ali ibn Tabataba Al-Fakhri: On the Systems or'
Govemment and the Moslem Dynasties. trans. C.E.I.Whitting, Hyperion
Press, Westport, Conn, 1981.
776
al-Tabari, Abu |afar Muhammad ibn larfu, Annales. ed. and trans. M. ]. de
Goeje, et. al. Leiden, L3 vols.
Secondarv Sources:
Ahsan, A.A. Social Life Under the Abbasids. Longman Group Ltd. London and
N.Y. 1979.
Allen, W.E.D. History of the Georgian People, Routledge, Keegan and Paul,
London,1977.
Balmes, N.H. The B]¡zantine Empire. Oxford University Press, Oxford, 1925.
Baynes, N.H., and Moss, H. St. L. B. Byzantium, Claredon Press, Oxford, 7948.
Canard, M. "Byzantium and the Muslim World to the Middle of The Eleventh
Century," Cambridge Medieval History, v.4,?std ed. pt.1-, 1966, c.3,679-
735.
Diehl, C. Historv of the Bvzantine Empire, AMS Press, New York, 1967. (1'st
ed. 1919).
"Leo LII and The Isaurian Dynasty (717-802)," CMH, Vol.4, Lst ed. 7936,
c. I (1-26).
Gero, S. Byzantine Iconoclasm Du¡ing the Reign of Leo III: with Particular
Attention to the Oriental Sources. Corpus Scriptorum Christianorum
Orientalium, Louvain, 1973.
Gibb, H.A.R. "The Arab l-nvasion of Kashgar in715," Bulletin of the Societ)¡ of
Oriental Studies.ii, pt.iü, 467 -7 4.
Goitein, S.D. Studies in Islar"nic History and Institution, E.]. Brill, Leidery 1968.
Haldory ].F. and Keruredy, FI. "The Arab-Byzantine Frontier in the Eighth and
Ninth Centuries: Military Organizatton and Society in the Borderlands,"
Zbornik Radova Vizantinoloskog Instituta,l 9 ( 1980) 7 9 -71'6.
Hawting, G.P. The First Dl¡nasty of Islam, South Illinois University Press,
Carbondale,Il,1986.
Huxley,G.''ontheVitaofSaintStephentheYounger,''@
B)¡zantine Studies. 18 G97n 97-L08.
"On the Vita of Saint ]ohn of Gottria," Greek. Roman and B]¡zantine
Studies. t9 (1978) 16I-69.
]ones, A.H.M. The Later Roman Empire. Basil Blackwell Publ; Oxford, 1964.
Levy, R. 'Persia and the Arabs," The Legac]¿ of Persia. ed. A.J. Arberry,
Clarendon Press, Oxford, 1953, 60-89.
"Moslem Trade in the Mediterranean and the West " The Islamic World
andtherffest.622-1492. ed. A. Lewis, John Wiley and Sons,Inc. N.Y.
1970,30-47.
Sumner, G.V. " Phitlipicus, Anastasius tr, and Theodosius III," Greek. Roman
+and Bvzantine Studies, 17 (1976) 287-94.
Tisdall, W. The Sources of Islam, trans. Muir, Amarko Books Agency, New
Delhi, 1961,.
Tritle, L.A. 'Tatzates Flight and The Byzantine Arab Peace Treafy of 782,"
B]¡zantion. 40 (197n, 297 -300.
Wellhausery J. The Arab Kingdom and lts Fall, trans. Weir, Calcutta, 1927.
124